Chat with us, powered by LiveChat Research Paper (chapter 5 and 6) | excelpaper.org/
+1(978)310-4246 credencewriters@gmail.com
  

Behavior
Modification
Principles and Procedures

FIFTH EDITION

RAYMOND G. MILTENBERGER
University of South Florida

Australia • Brazil • Japan • Korea • Mexico • Singapore • Spain • United Kingdom • United States

This is an electronic version of the print textbook. Due to electronic rights restrictions,
some third party content may be suppressed. Editorial review has deemed that any suppressed
content does not materially affect the overall learning experience. The publisher reserves the right
to remove content from this title at any time if subsequent rights restrictions require it. For
valuable information on pricing, previous editions, changes to current editions, and alternate
formats, please visit www.cengage.com/highered to search by ISBN#, author, title, or keyword for
materials in your areas of interest.

Behavior Modification: Principles and
Procedures, Fifth Edition
Raymond G. Miltenberger

Acquisitions Editor: Timothy Matray

Development Editor: Philip Hovanessian

Assistant Editor: Sheli DeNola

Editorial Assistant: Laren Moody

Marketing Communications Manager:
Sean Foy

Senior Art Director: Pamela Galbreath

Print Buyer: Rebecca Cross

Production Service: PreMediaGlobal

Cover Designer: Kathy Heming

Cover Image: © Olaru Radian-Alexandru/
Shutterstock.com

© 2012, 2008 Wadsworth, Cengage Learning

ALL RIGHTS RESERVED. No part of this work covered by the copyright
herein may be reproduced, transmitted, stored, or used in any form or by
any means graphic, electronic, or mechanical, including but not limited to
photocopying, recording, scanning, digitizing, taping, Web distribution,
information networks, or information storage and retrieval systems, except
as permitted under Section 107 or 108 of the 1976 United States Copyright
Act, without the prior written permission of the publisher.

For product information and technology assistance, contact us at
Cengage Learning Customer & Sales Support, 1-800-354-9706.

For permission to use material from this text or product,
submit all requests online at www.cengage.com/permissions.

Further permissions questions can be e-mailed to
[email protected]

Library of Congress Control Number: 2011926940

ISBN-13: 978-1-111-30611-3

ISBN-10: 1-111-30611-7

Wadsworth
20 Davis Drive
Belmont, CA 94002-3098
USA

Cengage Learning is a leading provider of customized learning solutions
with office locations around the globe, including Singapore, the United
Kingdom, Australia, Mexico, Brazil, and Japan. Locate your local office at
international.cengage.com/region.

Cengage Learning products are represented in Canada by
Nelson Education, Ltd.

For your course and learning solutions, visit www.cengage.com

Purchase any of our products at your local college store or at our preferred
online store www.cengagebrain.com.

Printed in the United States of America
1 2 3 4 5 6 7 15 14 13 12 11

To my wife, Nasrin,

and my kids, Ryan, Roxanne, and Steven

ABOUT THE AUTHOR

Raymond G. Miltenberger received his Ph.D. in Clinical Psychology in 1985from Western Michigan University. He is currently a professor and director
of the Applied Behavior Analysis Masters Program at the University of South Flor-
ida. Dr. Miltenberger conducts applied behavior analysis research with his students
and publishes widely in the areas of sports and fitness, functional assessment and
treatment of behavioral disorders, and self-protection skills training. He utilizes
behavior modification in clinical work with children and individuals with intellec-
tual disability. In addition to spending time with his family, he enjoys running,
golf, baseball, and travel.

iv

BRIEF CONTENTS

One / Introduction to Behavior Modification 1

PART 1 Measurement of Behavior and Behavior Change
Two / Observing and Recording Behavior 17
Three / Graphing Behavior and Measuring Change 39

PART 2 Basic Principles
Four / Reinforcement 61
Five / Extinction 87
Six / Punishment 101
Seven / Stimulus Control: Discrimination and Generalization 123
Eight / Respondent Conditioning 141

PART 3 Procedures to Establish New Behavior
Nine / Shaping 159
Ten / Prompting and Transfer of Stimulus Control 177
Eleven / Chaining 197
Twelve / Behavioral Skills Training Procedures 217

PART 4 Procedures to Increase Desirable Behavior and Decrease
Undesirable Behavior
Thirteen / Understanding Problem Behaviors through Functional Assessment 237
Fourteen / Applying Extinction 265
Fifteen / Differential Reinforcement 287
Sixteen / Antecedent Control Procedures 315
Seventeen / Using Punishment: Time-Out and Response Cost 341
Eighteen / Positive Punishment Procedures and the Ethics of Punishment 361
Nineteen / Promoting Generalization 379

PART 5 Other Behavior Change Procedures
Twenty / Self-Management 399
Twenty-One / Habit Reversal Procedures 417
Twenty-Two / The Token Economy 431
Twenty-Three / Behavioral Contracts 451
Twenty-Four / Fear and Anxiety Reduction Procedures 469
Twenty-Five / Cognitive Behavior Modification 489

v

CONTENTS

Preface xvi

Chapter One Introduction to Behavior Modification 1

Defining Human Behavior 2

Examples of Behavior 4

Defining Behavior Modification 5

Characteristics of Behavior Modification 5

Historical Roots of Behavior Modification 7

Major Figures 7
Early Behavior Modification Researchers 9
Major Publications and Events 9

Areas of Application 10

Developmental Disabilities 10
Mental Illness 11
Education and Special Education 11
Rehabilitation 12
Community Psychology 12
Clinical Psychology 12
Business, Industry, and Human Services 12
Self-Management 13

Child Behavior Management 13
Prevention 13
Sports Performance 13
Health-Related Behaviors 13
Gerontology 14

The Structure of This Textbook 14

Measurement of Behavior and Behavior
Change 14

Basic Principles of Behavior 14
Procedures to Establish New Behaviors 14
Procedures to Increase Desirable Behaviors and

Decrease Undesirable Behaviors 15
Other Behavior Change Procedures 15

Chapter Summary 15

Key Terms 16

Practice Test 16

PART 1 Measurement of Behavior and Behavior Change

Chapter Two Observing and Recording Behavior 17

Direct and Indirect Assessment 17

Defining the Target Behavior 19

The Logistics of Recording 21

The Observer 21
When and Where to Record 21

Choosing a Recording Method 23

Continuous Recording 23

Percentage of Opportunities 26
Product Recording 27
Interval Recording 27
Time Sample Recording 28

Choosing a Recording Instrument 29

Reactivity 32

Interobserver Agreement 33

vi

Chapter Summary 35

Key Terms 36

Practice Test 36

Applications 37

Misapplications 38

Chapter Three Graphing Behavior and Measuring Change 39

Components of a Graph 40

Graphing Behavioral Data 43

Graphing Data from Different Recording
Procedures 46

Research Designs 47

A-B Design 47
A-B-A-B Reversal Design 48
Multiple-Baseline Design 49

Alternating-Treatments Design 53
Changing-Criterion Design 55

Chapter Summary 57

Key Terms 57

Practice Test 58

Applications 58

Misapplications 59

PART 2 Basic Principles

Chapter Four Reinforcement 61

Defining Reinforcement 63

Positive and Negative Reinforcement 66

Social versus Automatic Reinforcement 68

Escape and Avoidance Behaviors 69

Conditioned and Unconditioned Reinforces 70

Factors That Influence the Effectiveness of
Reinforcement 72

Immediacy 72
Contingency 72
Motivating Operations 73
Individual Differences 75
Magnitude 75

Schedules of Reinforcement 76

Fixed Ratio 78
Variable Ratio 78
Fixed Interval 79
Variable Interval 80

Reinforcing Different Dimensions of
Behavior 81

Concurrent Schedules of Reinforcement 82

Chapter Summary 82

Key Terms 83

Practice Test 83

Appendix A 84

Appendix B 85

Chapter Five Extinction 87

Defining Extinction 88

Extinction Burst 90

Spontaneous Recovery 93

Procedural Variations of Extinction 93

A Common Misconception about
Extinction 96

Contents vii

Factors That Influence Extinction 96

Chapter Summary 98

Key Terms 98

Practice Test 99

Appendix A 99

Chapter Six Punishment 101

Defining Punishment 101

A Common Misconception about
Punishment 104

Positive and Negative Punishment 105

Unconditioned and Conditioned
Punishers 110

Contrasting Reinforcement and
Punishment 111

Factors That Influence the Effectiveness of
Punishment 114

Immediacy 114
Contingency 114
Motivating Operations 115
Are these examples of AOs or EOs? 115
Is this an example of an EO or an AO? 115

Individual Differences and Magnitude of the
Punisher 116

Problems with Punishment 117

Emotional Reactions to Punishment 117
Escape and Avoidance 117
Negative Reinforcement for the Use of

Punishment 117
Describe how the use of punishment may be

negatively reinforcing. 118
Punishment and Modeling 118
Ethical Issues 119

Chapter Summary 119

Key Terms 119

Practice Test 120

Appendix A 121

Chapter Seven Stimulus Control: Discrimination and
Generalization 123

Examples of Stimulus Control 124

Defining Stimulus Control 125

Developing Stimulus Control: Stimulus
Discrimination Training 126

Discrimination Training in the Laboratory 127
Developing Reading and Spelling with
Discrimination Training 128

Stimulus Discrimination Training and
Punishment 129

The Three-Term Contingency 130

Stimulus Control Research 130

Generalization 132

Examples of Generalization 133

Chapter Summary 137

Key Terms 138

Practice Test 138

Appendix A 138

Chapter Eight Respondent Conditioning 141

Examples of Respondent Conditioning 141

Defining Respondent Conditioning 142

Timing of the Neutral Stimulus and
Unconditioned Stimulus 145

viii Contents

Higher-Order Conditioning 146

Conditioned Emotional Responses 147

Extinction of Conditioned Responses 149

Spontaneous Recovery 149

Discrimination and Generalization of
Respondent Behavior 150

Factors That Influence Respondent
Conditioning 150

The Nature of the Unconditioned Stimulus and
Conditioned Stimulus 151

The Temporal Relationship between the
Conditioned Stimulus and Unconditioned
Stimulus 151

Contingency between the Conditioned Stimulus
and Unconditioned Stimulus 151

The Number of Pairings 151
Previous Exposure to the Conditioned
Stimulus 152

Distinguishing between Operant and
Respondent Conditioning 152

Respondent Conditioning and Behavior
Modification 155

Chapter Summary 156

Key Terms 156

Practice Test 156

PART 3 Procedures to Establish New Behavior

Chapter Nine Shaping 159

An Example of Shaping: Teaching a Child to
Talk 159

Defining Shaping 160

Applications of Shaping 162

Getting Mrs. F to Walk Again 162
Getting Mrs. S to Increase the Time between
Bathroom Visits 162

Research on Shaping 164

How to Use Shaping 168

Shaping of Problem Behaviors 170

Chapter Summary 173

Key Terms 174

Practice Test 174

Applications 174

Misapplications 175

Chapter Ten Prompting and Transfer of Stimulus Control 177

An Example of Prompting and Fading: Teaching
Little Leaguers to Hit the Ball 177

What Is Prompting? 179

What Is Fading? 180

Types of Prompts 182

Response Prompts 182
Stimulus Prompts 184

Transfer of Stimulus Control 185

Prompt Fading 186
Prompt Delay 188

Stimulus Fading 189

How to Use Prompting and Transfer of Stimulus
Control 191

Prompting and Transfer of Stimulus Control in
Autism Treatment 193

Chapter Summary 193

Key Terms 194

Practice Test 194

Applications 195

Misapplications 195

Contents ix

Chapter Eleven Chaining 197

Examples of Behavioral Chains 197

Analyzing Stimulus–Response Chains 198

Task Analysis 199

Backward Chaining 202

Forward Chaining 205

Total Task Presentation 206

Other Strategies for Teaching Behavioral
Chains 210

Written Task Analysis 210

Picture Prompts 210
Video Modeling 211
Self-Instructions 212

How to Use Chaining Procedures 214

Chapter Summary 215

Key Terms 215

Practice Test 215

Applications 216

Misapplications 216

Chapter Twelve Behavioral Skills Training Procedures 217

Examples of Behavioral Skills Training
Procedures 217

Teaching Marcia to Say “No” to the
Professors 217

Teaching Children to Protect Themselves from
Abduction 218

Components of the Behavioral Skills Training
Procedure 218

Modeling 218
Instructions 221
Rehearsal 222
Feedback 223

Enhancing Generalization after Behavioral Skills
Training 223

In Situ Assessment 224

In Situ Training 225

Behavioral Skills Training and the Three-Term
Contingency 225

Behavioral Skills Training in Groups 226

Applications of Behavioral Skills Training
Procedures 227

How to Use Behavioral Skills Training
Procedures 231

Chapter Summary 233

Key Terms 233

Practice Test 233

Applications 234

Misapplications 234

PART 4 Procedures to Increase Desirable Behavior and
Decrease Undesirable Behavior

Chapter Thirteen Understanding Problem Behaviors through
Functional Assessment 237

Examples of Functional Assessment 237

Jacob 237
Anna 239

Defining Functional Assessment 240

Functions of Problem Behaviors 241

Social Positive Reinforcement 241
Social Negative Reinforcement 241
Automatic Positive Reinforcement 242

x Contents

Automatic Negative Reinforcement 242

Functional Assessment Methods 242

Indirect Methods 243
Direct Observation Methods 245
Experimental Methods (Functional
Analysis) 250

Functional Analysis Research 254

Conducting a Functional Assessment 257

Functional Interventions 259

Chapter Summary 260

Key Terms 260

Practice Test 260

Applications 261

Misapplications 263

Chapter Fourteen Applying Extinction 265

The Case of Willy 265

Using Extinction to Decrease a Problem
Behavior 268

Collecting Data to Assess Treatment
Effects 268

Identifying the Reinforcer for the Problem
Behavior through Functional
Assessment 269

Eliminating the Reinforcer after Each Instance of
the Problem Behavior 269

Taking Account of the Schedule of
Reinforcement before Extinction 276

Reinforcing Alternative Behaviors 277

Promoting Generalization and
Maintenance 278

Research Evaluating the Use of Extinction 279

Chapter Summary 282

Key Terms 282

Practice Test 283

Applications 283

Misapplications 284

Appendix A 285

Appendix B 285

Chapter Fifteen Differential Reinforcement 287

Differential Reinforcement of Alternative
Behavior 287

Getting Mrs. Williams to Be Positive 287
When to Use DRA 289
How to Use DRA 290
Using Differential Negative Reinforcement of
Alternative Behaviors 293

Variations of DRA 295
Research on DRA 296

Differential Reinforcement of Other
Behavior 297

Defining DRO 299
Research Evaluating DRO Procedures 302

Differential Reinforcement of Low Rates of
Responding 305

Defining DRL 306
Variations of DRL 306
How are DRO and spaced-responding DRL
different? 307

Implementing DRL Procedures 307
Research Evaluating DRL Procedures 309

Chapter Summary 311

Key Terms 312

Practice Test 312

Applications 313

Misapplications 313

Contents xi

Chapter Sixteen Antecedent Control Procedures 315

Examples of Antecedent Control 315

Getting Marianne to Study More 315
Getting Cal to Eat Right 316

Defining Antecedent Control Procedures 317

Presenting the Discriminative Stimulus (SD) or
Cues for the Desired Behavior 317

Arranging Establishing Operations for the
Desirable Behavior 319

Decreasing Response Effort for the Desirable
Behavior 320

Removing the Discriminative Stimulus or Cues
for Undesirable Behaviors 322

Presenting Abolishing Operations for Undesirable
Behaviors 323

Increasing the Response Effort for Undesirable
Behaviors 324

Research on Antecedent Control
Strategies 325

Manipulating Discriminative Stimuli 326
Manipulating Response Effort 328
Manipulating Motivating Operations 329

Using Antecedent Control Strategies 334

Analysis of the Three-Term Contingency for the
Desirable Behavior 334

Analysis of the Three-Term Contingency for the
Undesirable Behavior 335

Functional Interventions for Problem
Behaviors 335

Chapter Summary 336

Key Terms 336

Practice Test 336

Applications 337

Misapplications 338

Chapter Seventeen Using Punishment: Time-out and
Response Cost 341

Time-out 342

Types of Time-out 344
Using Reinforcement with Time-out 344
Considerations in Using Time-out 345
Research Evaluating Time-out Procedures 348

Response Cost 350

Defining Response Cost 352
Using Differential Reinforcement with Response
Cost 352

Comparing Response Cost, Time-out, and
Extinction 352

Considerations in Using Response Cost 353
Research Evaluating Response Cost

Procedures 355

Chapter Summary 357

Key Terms 357

Practice Test 357

Applications 358

Misapplications 358

Chapter Eighteen Positive Punishment Procedures and the Ethics
of Punishment 361

Application of Aversive Activities 361

Overcorrection 363
Contingent Exercise 364
Guided Compliance 366

xii Contents

Physical Restraint 367
Cautions in the Application of Aversive
Activities 368

Application of Aversive Stimulation 368

Positive Punishment: Treatment of Last
Resort 371

Considerations in Using Positive
Punishment 372

The Ethics of Punishment 373

Informed Consent 374
Alternative Treatments 374
Recipient Safety 374

Problem Severity 374
Implementation Guidelines 374
Training and Supervision 375
Peer Review 375
Accountability: Preventing Misuse and
Overuse 375

Chapter Summary 375

Key Terms 376

Practice Test 376

Applications 377

Misapplications 377

Chapter Nineteen Promoting Generalization 379

Examples of Generalization Programming 379

Defining Generalization 380

Strategies for Promoting Generalization of
Behavior Change 381

Reinforcing Occurrences of Generalization 381
Training Skills That Contact Natural
Contingencies of Reinforcement 382

Modifying Contingencies of Reinforcement
and Punishment in the Natural
Environment 383

Incorporating a Variety of Relevant Stimulus
Situations in Training 386

Incorporating Common Stimuli 387

Teaching a Range of Functionally Equivalent
Responses 389

Incorporating Self-Generated Mediators of
Generalization 389

Implementing Strategies to Promote
Generalization 391

Promoting Generalized Reductions in Problem
Behaviors 392

Chapter Summary 394

Key Terms 395

Practice Test 395

Applications 396

Misapplications 396

PART 5 Other Behavior Change Procedures

Chapter Twenty Self-Management 399

Examples of Self-Management 399

Getting Murray to Run Regularly 399
Getting Annette to Clean up Her Mess 400

Defining Self-Management Problems 402

Defining Self-Management 404

Types of Self-Management Strategies 404

Goal-Setting and Self-Monitoring 404

Antecedent Manipulations 405
Behavioral Contracting 406
Arranging Reinforcers and Punishers 407
Social Support 408
Self-Instructions and Self-Praise 408

Steps in a Self-Management Plan 409

Clinical Problems 412

Contents xiii

Chapter Summary 413

Key Terms 413

Practice Test 414

Applications 414

Misapplications 415

Chapter Twenty-One Habit Reversal Procedures 417

Examples of Habit Behaviors 417

Defining Habit Behaviors 418

Nervous Habits 418
Motor and Vocal Tics 419
Stuttering 420

Habit Reversal Procedures 420

Applications of Habit Reversal 421

Nervous Habits 421
Motor and Vocal Tics 422
Stuttering 424

Why Do Habit Reversal Procedures Work? 426

Other Treatment Procedures for Habit
Disorders 427

Chapter Summary 428

Key Terms 429

Practice Test 429

Applications 429

Misapplications 430

Chapter Twenty-Two The Token Economy 431

Rehabilitating Sammy 431

Defining a Token Economy 432

Implementing a Token Economy 434

Defining the Target Behaviors 434
Identifying the Items to Use as Tokens 434
Identifying Backup Reinforcers 435
Deciding on the Appropriate Schedule of
Reinforcement 436

Establishing the Token Exchange Rate 437
Establishing the Time and Place for Exchanging
Tokens 437

Deciding Whether to Use Response Cost 438

Staff Training and Management 439

Practical Considerations 439

Applications of a Token Economy 441

Advantages and Disadvantages of a Token
Economy 447

Chapter Summary 447

Key Terms 448

Practice Test 448

Applications 448

Misapplications 449

Chapter Twenty-Three Behavioral Contracts 451

Examples of Behavioral Contracting 451

Getting Steve to Complete His
Dissertation 451

Helping Dan and His Parents Get Along
Better 453

Defining the Behavioral Contract 453

Components of a Behavioral Contract 454

Types of Behavioral Contracts 456

One-Party Contracts 457
Two-Party Contracts 457

Negotiating a Behavioral Contract 459

xiv Contents

Why Do Behavioral Contracts Influence
Behavior? 460

Applications of Behavioral Contracts 461

Chapter Summary 465

Key Terms 465

Practice Test 465

Applications 466

Misapplications 466

Chapter Twenty-Four Fear and Anxiety Reduction Procedures 469

Examples of Fear and Anxiety Reduction 469

Overcoming Trisha’s Fear of Public
Speaking 469

Overcoming Allison’s Fear of Spiders 470

Defining Fear and Anxiety Problems 471

Procedures to Reduce Fear and Anxiety 474

Relaxation Training 474
Systematic Desensitization 479
In Vivo Desensitization 482

Advantages and Disadvantages of Systematic
and In Vivo Desensitization 484

Other Treatments for Fears 485

Clinical Problems 486

Chapter Summary 486

Key Terms 487

Practice Test 487

Applications 487

Misapplications 488

Chapter Twenty-Five Cognitive Behavior Modification 489

Examples of Cognitive Behavior
Modification 489

Helping Deon Control His Anger 489
Helping Claire Pay Attention in Class 491

Defining Cognitive Behavior Modification 492

Defining Cognitive Behavior 492
Functions of Cognitive Behavior 494

Cognitive Behavior Modification
Procedures 494

Cognitive Restructuring 495

Cognitive Coping Skills Training 499
Acceptance-Based Therapies 501

Clinical Problems 502

Chapter Summary 502

Key Terms 502

Practice Test 503

Applications 503

Misapplications 504

Glossary 506

References 517

Name Index 541

Subject Index 549

Quizzes Q1

Contents xv

PREFACE

Iam gratified that the first four editions of Behavior Modification: Principles andProcedures received positive reviews from students and professors. The fifth
edition has kept the positive features of the first four editions, has been revised to
address the suggestions of reviewers, and has been updated to reflect the latest
research in behavior modification.

The goal of this fifth edition (as with the earlier editions) is to describe
basic principles of behavior so that the student learns how environmental events
influence human behavior and to describe behavior modification procedures so
that the student learns the strategies by which human behavior may be changed.
The text is divided into 25 relatively short chapters, each of which covers a man-
ageable amount of information (for example, one principle or procedure). This
text can be used in a standard one-semester course in behavior modification,
applied behavior analysis, behavior management, or behavior change.

The material in the text is discussed at an introductory level so that it may be
understood by students with no prior knowledge of the subject. This text is
intended for undergraduate students or beginning graduate students. It would also
be valuable for individuals working in human services, education, or rehabilitation
who must use behavior modification procedures to manage the behavior of the
individuals in their care.

I have made a concerted effort in this text to be gender neutral. When dis-
cussing case examples, I include males and females about equally as often.

Features of the Text Continued from the
First Four Editions

The following features of the text are intended to help the reader learn easily.

Organization of the Text Following a general introduction to the field in
Chapter 1, Chapters 2 and 3 present information on behavior recording, graphing,
and measuring change. This information will be utilized in each subsequent
chapter. Next, Chapters 4–8 focus on the basic principles of operant and respon-
dent behavior. The application of these principles forms the subject of the remain-
ing 17 chapters. Procedures to establish new behaviors are described in Chapters
9–12, and procedures to increase desirable behaviors and decrease undesirable
behaviors are considered in Chapters 13–19. Finally, Chapters 20–25 present a
survey of other important behavior modification procedures.

Principles and Procedures The various procedures for changing behavior are
based on fundamental principles of behavior established in experimental

xvi

research over the last 80 years. In the belief that the student will better under-
stand the procedures after first learning the fundamental principles, the princi-
ples underlying operant and respondent behavior are reviewed in Chapters 4–8;
the application of the principles in the behavior modification procedures is
described in Chapters 9–25.

Examples from Everyday Life Each chapter uses a variety of real-life
examples—some relevant to college students, some chosen from the author’s
clinical experience—to bring the principles and procedures to life.

Examples from Research In addition, both classic studies and the most
up-to-date research on behavior modification principles and procedures are inte-
grated into the text.

Quizzes Accompany Each Chapter Three fill-in-the-blank quizzes with 10
questions are provided for each of the 25 chapters. The quizzes provide students
with further exercises for self-assessment of their knowledge of the chapters’ con-
tent. The quizzes are on perforated pages, which can be easily torn out so that
the instructor can have students hand the quizzes in as homework assignments or
have students take the quizzes in class.

Practice Tests Practice tests at the end of each chapter have short-answer essay
questions, complete with page numbers where the answers can be found.

Application Exercises At the end of each chapter where procedures are taught
(Chapters 2, 3, and 9–25), several application exercises are provided. In each exer-
cise, a real-life case is described and then the student is asked to apply the proce-
dure described in the chapter. These exercises give students an opportunity to
think about how the procedures are applied in real life.

Misapplication Exercises The application exercises are followed by misapplica-
tion exercises. In each one, a case example is provided, and the procedure from
the chapter is applied to the case in an incorrect or inappropriate manner. The
student is asked to analyze the case example and to describe what is wrong with
the application of the procedure in that case. These misapplication exercises
require the student to think critically about the application of the procedure.
Answers to Applications and Misapplications are in the Instructors Manual, mak-
ing them valuable tools for instructors as they assess their students’ abilities to
apply the information provided in the chapter.

Step-by-Step Approach In each chapter in which a particular behavior modifica-
tion procedure is taught, the implementation of the procedure is outlined in a
step-by-step fashion, for ease of comprehension.

Summary Boxes Periodically throughout the text, information from a chapter is
summarized in a box that has been set off from the text. These boxes are intended
to help the student organize the material in the chapter.

Preface xvii

Chapter Summaries Chapter summaries provide information that is consistent
with the opening questions in each chapter.

Examples for Self-Assessment In the early chapters on basic principles
(Chapters 4–7) there are tables with examples of the principle discussed within
that chapter. Later in the chapter (or in a subsequent chapter), the student is
directed to return to a specific table and, using the new information being
presented in the chapter, to analyze specific aspects of the examples provided in
that table.

Self-Assessment Questions At intervals throughout the text, students are pre-
sented with self-assessment questions. To answer these questions, students will
need to utilize the information already presented in the chapter. These questions
will help students assess their understanding of the material. In most cases,
answers are presented in the text immediately following the question.

Figures Most of the chapters include figures from the research literature to illus-
trate important principles or procedures. Students must use information from ear-
lier chapters on behavior recording, graphing, and measuring change to analyze
the graphs.

Glossary At the end of the text is a glossary of the important behavior modifica-
tion terms used in the text. Each term is followed by a succinct and precise
definition.

Improved Test Bank The test bank includes multiple-choice questions, fill-in-
the-blank questions, true-false questions, and short-answer essay questions.

For Further Reading Each of the chapters includes a For Further Reading box.
In this feature, interesting articles that are relevant to the content of the chapter
are identified and briefly described. Citations for these articles have also been pro-
vided. These articles are from JABA (or JEAB), so they can be easily accessed
online by students. (The JABA website is http://seab.envmed.rochester.edu/jaba/.)
Instructors can assign these articles for extra credit or as reading assignments for
when more advanced students use the textbook.

List of Key Terms After each Chapter Summary section, there is a list of the
new terms that were used in the chapter. The list of key terms shows the page
number on which each term was introduced. Although these terms are all found
in the Glossary at the end of the text, having the new terms, and their page
numbers, listed at the end of each chapter will allow the student to have an easy
reference to the terms when reading the chapter or when studying for a test or
quiz.

xviii Preface

New Features in the Fifth Edition

Motivating Operations The term motivating operation is introduced in Chapter
4 and discussed thoughout the text (particularly in Chapter 4, Chapter 6, and
Chapter 16). Motivating operation is a more contemporary term that encompasses
establishing operations and abolishing operations. As such, the term abolishing
operation is also introduced in Chapter 4 and discussed in the text.

Reinforcer Assessment Procedures The process of choosing reinforcers is
discussed in Chapter 15 (Differential Reinforcement). The book now identifies
and describes three different approaches to reinforcer assessment established in
the literature: single stimulus assessment, paired stimulus assessment, and multiple
stimulus assessment procedures.

On Terms A new feature called “On Terms” has been added in a few chapters
(Chapters 4, 5, 6, and 8) to help students identify the correct ways to use various
behavioral terms. This feature is added because of the author’s experience with
mistakes in terminology students commonly make.

Interobserver Agreement The term “interobserver reliability” has been changed
to the more widely accepted term “interobserver agreement,” both in Chapter 2
and throughout the text. In addition, Chapter 2 discusses two additional strategies
for conducting interobserver agreement: occurrence only and nonoccurrence only
agreement.

Research Designs In Chapter 3, a discussion of variations of the ABAB reversal
design has been added, along with a discussion of the nonconcurrent multiple
baseline design, which is a variation of the multiple baseline across subjects
design.

Extinction In Chapter 14, the distinction between extinction applied to beha-
viors maintained by positive versus negative reinforcement has been clarified, the
term escape extinction has been introduced, and the potential difficulties that
may be encountered when implementing escape extinction are now discussed.

Functional Assessment A number of changes have been made to the chapter on
functional assessment (Chapter 13). For this edition, the author has: mentioned
that scatterplot is not an ABC recording method; highlighted the three methods
for conducting ABC direct observation methods with the use of bullets; clarified
the definition of a functional analysis; differentiated between an exploratory func-
tional analysis and a hypothesis testing functional analysis; clarified the role of test
and control conditions in a functional analysis; and finally, in the “conducting a
functional assessment” section, numbered the steps to make them more clear.

Quizzes at the End of the Book The three quizzes at the end of each chapter
are now all together at the end of the book, color tabbed for ease in locating, and

Preface xix

perforated for easy removal and handing-in during class. The pages in the book
now are no longer perforated except for the quizzes at the end of the book.

Other New Features
â–  Added discussion of social validity and the use of technology in data collec-

tion (Chapter 2)
â–  Added discussion of recent publications on the use of Excel for graphing

(Chapter 2)
â–  Clarified the distinction between positive and negative reinforcement

(Chapter 4) and positive and negative punishment (Chapter 6)
â–  Added brief discussion of prompting and fading for use in autism

(Chapter 10)
â–  Added brief section on video modeling (Chapter 11)
â–  Added brief section on in situ assessments (Chapter 12)
â–  Moved discussion of functional interventions from the end of Chapter 16 to

the end of Chapter 13, immediately following discussion of functional assessment
procedures

â–  Added over 50 new references throughout the text
â–  Changed the term mental retardation to the more acceptable and contem-

porary term intellectual disability throughout the text
â–  Added a dozen new terms to the text and included them in the glossary

Acknowledgments

I want to thank the anonymous reviewers for their constructive comments on this
manuscript and the first four editions: Robert W. Allan, Lafayette College; Viviette
Allen, Fayetteville State University; Cynthia Anderson, West Virginia University;
Jennifer Austin, Florida State University; Charles Blose, MacMurry College;
Kristine Brady, California School of Professional Psychology; James Carr, Western
Michigan University; Carl Cheney, Utah State University; Darlene Crone-Todd,
Delta State University; Paula Davis, Southern Illinois University; Richard N. Feil,
Mansfield University; Deirdre Beebe Fitzgerald, Eastern Connecticut State
University; Stephan Flanagan, The University of North Carolina at Chapel Hill;
Roger Harnish, Rochester Institute of Technology; Gerald Harris, The University
of Houston; Robert Heffer, Texas A&M University; Stephen W. Holborn, Univer-
sity of Manitoba; Dorothea Lerman, Louisiana State University; Tom Lombardo,
University of Mississippi; John Malouff, Nova Southern Eastern University;
Guenn Martin, Cumberland University; Kay McIntyre, University of Missouri–St.
Louis; Ronald Miller, Brigham Young University—Hawaii; Robert W. Montgom-
ery, Georgia State University; Charles S. Peyser, University of the South; Brady
Phelps, South Dakota State University; Joseph J. Plaud, University of North
Dakota; Robyn Rogers, Southwest Texas State University; Johannes Rojahn,
George Mason University; Paul Romanowich, Mesa College; Alison Thomas-
Cottingham, Rider University; J. Kevin Thompson, University of Southern
Florida; Bruce Thyer, University of Georgia; James T. Todd, Eastern Michigan
University; Sharon Van Leer, Delaware State University; Timothy Vollmer,

xx Preface

University of Florida; Robert W. Wildblood, Northern Virginia Community
College; Kenneth N. Wildman, Ohio Northern University; Douglas Woods,
University of Wisconsin–-Milwaukee; and Todd Zakrajsek, Southern Oregon State
College. I especially want to thank Marianne Taflinger, senior editor at Wadsworth,
for her guidance and support throughout the development of the first four editions.

For the Behavior Modification Student

To get the most out of this text and out of your behavior modification course, you
are encouraged to consider the following recommendations.

1. Read the assigned chapters before the class meeting at which the chapter is
to be discussed. You will benefit more from the class if you have first read
the material.

2. Answer each of the self-assessment questions in the chapter to see if you
understand the material just covered.

3. Answer the practice test questions at the end of each chapter. If you can
answer each question, you know that you understand the material in the
chapter.

4. Complete the end-of-chapter quizzes to assess your knowledge of the
chapter content (unless your professor plans to use the quizzes in class).

5. Complete the application and misapplication exercises at the end of the
procedure chapters. In that way, you will understand the material in the
chapter well enough to apply it or to identify how it is applied incorrectly.

6. The best way to study for a test is to test yourself. After reading and reread-
ing the chapter and your class notes, test yourself in the following ways.

â–  Look at key terms in the chapter and see if you can define them without
looking at the definitions in the text.

â–  Look at each practice test question at the end of the chapter and see if
you can give the correct answer without looking up the answer in the
text or in your notes.

â–  Come up with novel examples of each principle or procedure in the
chapter.

â–  Make flash cards with a term or question on one side and the definition
of the term or the answer to the question on the other side. While study-
ing, look at the term (or question) on one side of the card and then read
the definition (or answer) on the other. As you study, you will find that
you need to turn the cards over less and less often. Once you can supply
the answer or definition on the back of the card without looking, you’ll
know that you understand the material.

â–  Always study in a location that is reasonably free from distractions or
interruptions.

â–  Always begin studying for a test at least a few days in advance. Give your-
self more days to study as more chapters are included on the test.

Preface xxi

The following websites provide a range of valuable information about different aspects of behavior modification or applied
behavior analysis.

http://seab.envmed.rochester.edu/jaba/
http://seab.envmed.rochester.edu/jeab/
http://www.abainternational.org
http://www.auburn.edu/~newlamc/apa_div25/

Home

http://fabaworld.org
http://www.calaba.org/
http://www.unt.edu/behv/txaba/
http://www.babat.org/
http://www.behavior-analyst-online.org/index.html

A-Homepage

Home

BACB

Journal of Applied Behavior Analysis
Journal of the Experimental Analysis of Behavior
The Association for Behavior Analysis
APA Division 25 (Behavior Analysis)
Association for Behavioral and Cognitive Therapy
Florida Association for Behavior Analysis
California Association for Behavior Analysis
Texas Association for Behavior Analysis
Berkshire Association for Behavior Analysis and Therapy
The Behavior Analyst Online
Cambridge Center for Behavioral Studies
B.F. Skinner Foundation
Behavior Analysis Certification Board

Raymond G. Miltenberger

xxii Preface

Chapter One

Introduction to Behavior Modification

In this textbook you will learn about behavior modification, the principles andprocedures used to understand and change human behavior. Behavior modifica-
tion procedures come in many forms. Consider the following examples.

Ted and Jane were having some difficulties in their marriage because of fre-
quent arguments. Their marriage counselor arranged a behavioral contract with
them in which they agreed to do several nice things for each other every day. As
a result of this contract, their positive interactions increased and their negative
interactions (arguments) decreased.

Karen pulled her hair incessantly; as a result, she created a bald spot on
the top of her head. Although she was embarrassed by the bald spot, which

measured 1 inch in diameter, she continued to pull her hair.
Her psychologist implemented a treatment in which Karen was
to engage in a competing activity with her hands (e.g., needle-
point) each time she started to pull her hair or had the urge
to pull. Over time, the hair-pulling stopped and her hair grew
back in.

Francisco was gaining a lot of weight and decided to do some-
thing about it. He joined a weight loss group. At each group meet-
ing, Francisco deposited a sum of money, set a goal for daily
exercise, and earned points for meeting his exercise goals each
week. If he earned a specified number of points, he got his deposit
back. If he did not earn enough points, he lost part of his deposit

money. Francisco began to exercise regularly and lost weight as a result of his par-
ticipation in the group.

The residents of Cincinnati were making thousands of unnecessary directory
assistance calls per day. These calls were clogging up the phone lines and costing
the phone company money. The company instituted a charge for each directory
assistance call, and the number of calls decreased dramatically.

You will notice that each of these examples focuses on some aspect of human
behavior and describes ways to change the behavior. Because behavior modifica-
tion focuses on behavior and behavior change, it is appropriate to begin with a dis-
cussion of behavior.

â–  How is human behavior defined?

â–  What are the defining features of
behavior modification?

â–  What are the historical roots of
behavior modification?

â–  In what ways has behavior
modification improved people’s
lives?

1

Defining Human Behavior

Human behavior is the subject matter of behavior modification. Behavior is what
people do and say. The characteristics that define behavior are as follows.

■ Behavior involves a person’s actions (what people do or say); it is described
with action verbs. Behavior is not a static characteristic of the person. If you say
that a person is angry, you have not identified the person’s behavior; you have sim-
ply labeled the person. If you identify what the person says or does when angry,
then you have identified behavior. For example, “Jennifer screamed at her
mother, ran upstairs, and slammed the door to her room.” This is a description of
behavior that might be labeled as anger.

â–  Behaviors have one or more dimensions that can be measured. You can
measure the frequency of a behavior; that is, you can count the number of times
a behavior occurs (e.g., Shane bit his fingernails 12 times in the class period). You
can measure the duration of a behavior, or the time from when an instance of the
behavior starts until it stops (e.g., Rita jogged for 25 minutes). You can measure
the intensity of a behavior or the physical force involved in the behavior (e.g.,
Garth bench pressed 220 pounds). You can measure the speed of behavior, or
the latency from some event to the start of a behavior. Frequency, duration, inten-
sity, and latency are all physical dimensions of a behavior.

â–  Behaviors can be observed, described, and recorded by others or by the per-
son engaging in the behavior. Because a behavior is an action, its occurrence can
be observed. People can see the behavior (or detect it through one of the senses)
when it occurs. Because it is observable, the person who sees the behavior can
describe it and record its occurrence. (See Chapter 2 for a description of methods
for recording behavior.)

â–  Behaviors have an impact on the environment, including the physical or the
social environment (other people and ourselves). Because a behavior is an action
that involves movement through space and time (Johnston & Pennypacker, 1981),
the occurrence of a behavior has some effect on the environment in which it occurs.
Sometimes the effect on the environment is obvious. You turn the light switch, and
the light goes on (an effect on the physical environment). You raise your hand in
class, and your professor calls on you (an effect on other people). You recite a
phone number from a web site, and you are more likely to remember it and to dial
the correct number (an effect on yourself). Sometimes the effect of a behavior on the
environment is not obvious. Sometimes it has an effect only on the person who
engages in the behavior. However, all human behavior operates on the physical or
social environment in some way, regardless of whether we are aware of its impact.

â–  Behavior is lawful; that is, its occurrence is systematically influenced by envi-
ronmental events. Basic behavioral principles describe the functional relationships
between our behavior and environmental events. These principles describe how
our behavior is influenced by, or occurs as a function of, environmental events (see
Chapters 4–8). These basic behavioral principles are the building blocks of behavior
modification procedures. Once you understand the environmental events that cause
behaviors to occur, you can change the events in the environment to alter behavior.

Consider the graph in Figure 1-1, which shows the disruptive behavior of
a child with autism in the classroom. When the child receives high levels of

2 Chapter 1

attention from the teacher, his disruptive behavior rarely occurs. When the child
receives low levels of attention from the teacher, his disruptive behavior occurs
more frequently. We conclude that the disruptive behavior is functionally related
to the level of teacher attention.

â–  Behaviors may be overt or covert. Most often, behavior modification proce-
dures are used to understand and change overt behaviors. An overt behavior is an
action that can be observed and recorded by a person other than the one engaging
in the behavior. However, some behaviors are covert. Covert behaviors, also
called private events (Skinner, 1974), are not observable to others. For example,
thinking is a covert behavior; it cannot be observed and recorded by another per-
son. Thinking can be observed only by the person engaging in the behavior. The
field of behavior modification focuses primarily on overt or observable behaviors,
as does this textbook. However, Chapters 8, 24, and 25 discuss covert behaviors
and behavior modification procedures applied to them.

Characteristics of Behavior

Behavior is what people do and say.

Behaviors have one or more dimensions.

Behaviors can be observed, described, and recorded.

Behaviors have an impact on the environment.

Behavior is lawful.

Behaviors may be overt and covert.

FIGURE 1-1 This graph, adapted from a study by Durand and Carr (1992), shows the influence of teacher atten-
tion on the disruptive behavior (defined as pushing away task materials; loud screaming, whining, or
crying; and hitting or knocking over objects) of a young boy (Paul) in a special education classroom.
The graph shows that disruptive behavior does not occur when Paul receives frequent teacher
attention (High ATT). However, when Paul receives teacher attention infrequently (Low ATT), he
engages in disruptive behavior about 50% of the time. This graph shows the functional relationship
between the teacher’s attention and Paul’s disruptive behavior (From Durand, V. M., & Carr, E. G.
[1992]. An analysis of maintenance following functional communication training. Journal of Applied
Behavior Analysis, 25, 777–794. Copyright © 1992 University of Kansas Press. Reprinted by per-
mission of the author.)

Introduction to Behavior Modification 3

Examples of Behavior

Now let’s illustrate the defining characteristics of behavior with some examples.
The following examples include both common behaviors and problematic beha-
viors for which behavior modification procedures might be used.

Martha sits at her computer and types an e-mail to her parents.

This is behavior because pressing the keys on the keyboard while typing is an
action, has physical dimensions (frequency of pressing keys, duration of typing), is
observable and measurable, has an impact on the environment (produces letters
on the screen), and is lawful (occurs because of previous learning that pressing
the keys produces letters on the screen).

Mandy lies in her crib and cries loudly. Her mother then picks her up and feeds her.

This behavior has all five of the characteristics described in the previous example
(an action that has physical dimensions, is observable by others, produces an effect
on the environment, and is lawful). One difference is that the effect of crying is on
the social environment; her mother responds to her crying by picking her up and
feeding her. Each time it has occurred in the past, crying has resulted in her mother
feeding her, so the crying continues to occur when Mandy is hungry. There is a
functional relationship between the crying and the mother’s behavior of feeding her.

Jerry’s paper for his behavior modification class is a week late. Jerry gives the paper
to his professor and lies, saying that it is late because he had to go home to see his
sick grandmother. The professor then accepts the paper without any penalty. Jerry
also missed his history test. He tells his history professor he missed the test because
of his sick grandmother. The professor lets him take the test a week late.

Jerry’s behavior—lying about his visit to his sick grandmother—has all five
characteristics of a behavior. It is an action (something he said) that occurred
twice (frequency), was observed by his professors, and resulted in an effect on his
social environment (his professors let him take a test late and hand in a paper late
with no penalty); it is lawful because there is a functional relationship between
the behavior (lying) and the outcome (getting away with late papers or tests).

Samantha is a 6-year-old with an intellectual disability who attends special educa-
tion classes. When the teacher is helping other students and not paying attention to
Samantha, Samantha cries and bangs her head on the table or floor. Whenever
Samantha bangs her head, the teacher stops what she is doing and picks Samantha
up and comforts her. She tells Samantha to calm down, assures her that everything
is all right, gives her a hug, and often lets Samantha sit on her lap.

Identify each of the five characteristics of Samantha’s behavior.

Samantha’s head banging is a behavior. It is an action that she repeats a num-
ber of times each day. The teacher could observe and record the number of
occurrences each day. The head banging produces an effect on the social environ-
ment: The teacher provides attention each time the behavior occurs. Finally, the
behavior is lawful; it continues to occur because there is a functional relationship
between the head-banging behavior and the outcome of teacher attention.

4 Chapter 1

Defining Behavior Modification

Behavior modification is the field of psychology concerned with analyzing and
modifying human behavior.

â–  Analyzing means identifying the functional relationship between environ-
mental events and a particular behavior to understand the reasons for behavior or
to determine why a person behaved as he or she did.

â–  Modifying means developing and implementing procedures to help people
change their behavior. It involves altering environmental events so as to influence
behavior. Behavior modification procedures are developed by professionals and
used to change socially significant behaviors, with the goal of improving some
aspect of a person’s life. Following are some characteristics that define behavior
modification (Gambrill, 1977; Kazdin, 1994).

Characteristics of Behavior Modification
â–  Focus on behavior. Behavior modification procedures are designed to

change behavior, not a personal characteristic or trait. Therefore, behavior modifi-
cation de-emphasizes labeling. For example, behavior modification is not used to
change autism (a label); rather, behavior modification is used to change problem
behaviors exhibited by children with autism.

Behavioral excesses and deficits are targets for change with behavior modifica-
tion procedures. In behavior modification, the behavior to be modified is called
the target behavior. A behavioral excess is an undesirable target behavior the per-
son wants to decrease in frequency, duration, or intensity. Smoking is an example
of a behavioral excess. A behavioral deficit is a desirable target behavior the per-
son wants to increase in frequency, duration, or intensity. Exercise and studying
are possible examples of behavioral deficits.

â–  Procedures based on behavioral principles. Behavior modification is the
application of basic principles originally derived from experimental research with
laboratory animals (Skinner, 1938). The scientific study of behavior is called the
experimental analysis of behavior, or behavior analysis (Skinner, 1953b, 1966).
The scientific study of human behavior to help people change behavior in mean-
ingful ways is called applied behavior analysis (Baer, Wolf, & Risley, 1968,
1987). Behavior modification procedures are based on research in applied behav-
ior analysis that has been conducted for more than 50 years (Ullmann & Krasner,
1965; Ulrich, Stachnik, & Mabry, 1966).

â–  Emphasis on current environmental events. Behavior modification involves
assessing and modifying the current environmental events that are functionally
related to the behavior. Human behavior is controlled by events in the immedi-
ate environment, and the goal of behavior modification is to identify those
events. Once these controlling variables have been identified, they are altered
to modify the behavior. Successful behavior modification procedures alter the
functional relationships between the behavior and the controlling variables in
the environment to produce a desired change in the behavior. Sometimes labels
are mistakenly identified as the causes of behavior. For example, a person might

Introduction to Behavior Modification 5

say that a child with autism engages in problem behaviors (such as screaming,
hitting himself, refusal to follow instructions) because the child is autistic. In
other words, the person is suggesting that autism causes the child to engage in
the behavior. However, autism is simply a label that describes the pattern of
behaviors the child engages in. The label cannot be the cause of the behavior
because the label does not exist as a physical entity or event. The causes of the
behavior must be found in the environment (including the biology of the
child).

Behavior Modification and Applied Behavior Analysis

Behavior modification (as described in this textbook) and applied behavior analysis are two terms used to
identify virtually identical fields. Although research on the application of behavioral principles to help peo-
ple change their behavior (behavior modification) had been published since the late 1950s, the term
applied behavior analysis was introduced in 1968 in the first issue of Journal of Applied Behavior Analy-
sis with the publication of Baer, Wolf, and Risley’s article defining applied behavior analysis. In their arti-
cle, Baer et al. (1968) identified a number of characteristics of applied behavior analysis including: (a) a
focus on socially important behavior; (b) demonstration of functional relationships between environmental
events and behavior; (c) clear description of procedures; (d) connection to basic behavioral principles;
and (e) production of meaningful, generalizable, and long-lasting changes in behavior. These defining
features of applied behavior analysis also characterize the contemporary field of behavior modification
as described in this textbook.

â–  Precise description of behavior modification procedures (Baer et al., 1968).
Behavior modification procedures involve specific changes in environmental
events that are functionally related to the behavior. For the procedures to be
effective each time they are used, the specific changes in environmental events
must occur each time. By describing procedures precisely, researchers and other
professionals make it more likely that the procedures will be used correctly each
time.

â–  Treatment implemented by people in everyday life (Kazdin, 1994). Behavior
modification procedures are developed by professionals or paraprofessionals
trained in behavior modification. However, behavior modification procedures
often are implemented by people such as teachers, parents, job supervisors, or
others to help people change their behavior. People who implement behavior
modification procedures should do so only after sufficient training. Precise
descriptions of procedures and professional supervision make it more likely that
parents, teachers, and others will implement procedures correctly.

â–  Measurement of behavior change. One of the hallmarks of behavior modifi-
cation is its emphasis on measuring the behavior before and after intervention to
document the behavior change resulting from the behavior modification proce-
dures. In addition, ongoing assessment of the behavior is done well beyond the
point of intervention to determine whether the behavior change is maintained in
the long run. If a supervisor is using behavior modification procedures to increase
work productivity (to increase the number of units assembled each day), he or she
would record the workers’ behaviors for a period before implementing the proce-
dures. The supervisor would then implement the behavior modification proce-
dures and continue to record the behaviors. This recording would establish

6 Chapter 1

whether the number of units assembled increased. If the workers’ behaviors chan-
ged after the supervisor’s intervention, he or she would continue to record the
behavior for a further period. Such long-term observation would demonstrate
whether the workers continued to assemble units at the increased rate or whether
further intervention was necessary.

â–  De-emphasis on past events as causes of behavior. As stated earlier, behavior
modification places emphasis on recent environmental events as the causes of
behavior. However, knowledge of the past also provides useful information about
environmental events related to the current behavior. For example, previous learn-
ing experiences have been shown to influence current behavior. Therefore, under-
standing these learning experiences can be valuable in analyzing current behavior
and choosing behavior modification procedures. Although information on past
events is useful, knowledge of current controlling variables is most relevant to
developing effective behavior modification interventions because those variables,
unlike past events, can still be changed.

â–  Rejection of hypothetical underlying causes of behavior. Although some
fields of psychology, such as Freudian psychoanalytic approaches, might be inter-
ested in hypothesized underlying causes of behavior, such as an unresolved Oedi-
pus complex, behavior modification rejects such hypothetical explanations of
behavior. Skinner (1974) has called such explanations “explanatory fictions”
because they can never be proved or disproved, and thus are unscientific. These
supposed underlying causes can never be measured or manipulated to demon-
strate a functional relationship to the behavior they are intended to explain.

Characteristics of Behavior Modification

Focus on behavior

Based on behavioral principles

Emphasis on current environmental events

Precise description of procedures

Implemented by people in everyday life

Measurement of behavior change

De-emphasis on past events as causes of behavior

Rejection of hypothetical underlying causes of behavior

Historical Roots of Behavior Modification

A number of historical events contributed to the development of behavior modifi-
cation. Let’s briefly consider some important figures, publications, and organiza-
tions in the field.

Major Figures
Following are some of the major figures who were instrumental in developing
the scientific principles on which behavior modification is based (Figure 1-2)
(Michael, 1993a).

Introduction to Behavior Modification 7

Ivan P. Pavlov (1849–1936) Pavlov conducted experiments that uncovered the
basic processes of respondent conditioning (see Chapter 8). He demonstrated that
a reflex (salivation in response to food) could be conditioned to a neutral stimulus.
In his experiments, Pavlov presented the neutral stimulus (the sound of a metro-
nome) at the same time that he presented food to a dog. Later, the dog salivated
in response to the sound of the metronome alone. Pavlov called this a conditioned
reflex (Pavlov, 1927).

FIGURE 1-2 Four major figures who were instrumental in developing the scientific principles on which behavior
modification is based. Clockwise from top left: Ivan P. Pavlov, Edward L. Thorndike, B. F. Skinner,
John B. Watson. (Photo credits: SOV; Archives of the History of American Psychology, Center for the
History of Pychology-The University of Akron; Courtesy of the B. F. Skinner Foundation; Archives of the
History of American Psychology, Center for the History of Pychology-The University of Akron.)

8 Chapter 1

Edward L. Thorndike (1874–1949) Thorndike’s major contribution was the
description of the law of effect. The law of effect states that a behavior that pro-
duces a favorable effect on the environment is more likely to be repeated in the
future. In Thorndike’s famous experiment, he put a cat in a cage and set food out-
side the cage where the cat could see it. To open the cage door, the cat had to hit a
lever with its paw. Thorndike showed that the cat learned to hit the lever and open
the cage door. Each time it was put into the cage, the cat hit the lever more quickly
because that behavior—hitting the lever—produced a favorable effect on the envi-
ronment: It allowed the cat to reach the food (Thorndike, 1911).

John B. Watson (1878–1958) In the article “Psychology as the Behaviorist
Views It,” published in 1913, Watson asserted that observable behavior was the
proper subject matter of psychology, and that all behavior was controlled by envi-
ronmental events. In particular, Watson described a stimulus–response psychology
in which environmental events (stimuli) elicited responses. Watson started the
movement in psychology called behaviorism (Watson, 1913, 1924).

B. F. Skinner (1904–1990) Skinner expanded the field of behaviorism originally
described by Watson. Skinner explained the distinction between respondent condi-
tioning (the conditioned reflexes described by Pavlov and Watson) and operant condi-
tioning, in which the consequence of behavior controls the future occurrence of the
behavior (as in Thorndike’s law of effect). Skinner’s research elaborated the basic prin-
ciples of operant behavior (see Chapters 4–7). In addition to his laboratory research
demonstrating basic behavioral principles, Skinner wrote a number of books in
which he applied the principles of behavior analysis to human behavior (see later).
Skinner’s work is the foundation of behavior modification (Skinner, 1938, 1953a).

Early Behavior Modification Researchers
After Skinner laid out the principles of operant conditioning, researchers continued
to study operant behavior in the laboratory (Catania, 1968; Honig, 1966). In addi-
tion, in the 1950s, researchers began demonstrating behavioral principles and evalu-
ating behavior modification procedures with people. These early researchers studied
the behavior of children (Azrin & Lindsley, 1956; Baer, 1960; Bijou, 1957), adults
(Goldiamond, 1965; Verplanck, 1955; Wolpe, 1958), patients with mental illness
(Ayllon & Azrin, 1964; Ayllon & Michael, 1959), and individuals with intellectual dis-
abilities (Ferster, 1961; Fuller, 1949; Wolf, Risley, & Mees, 1964). Since the beginning
of behavior modification research with humans in the 1950s, thousands of studies have
established the effectiveness of behavior modification principles and procedures.

Major Publications and Events
A number of books heavily influenced the development of the behavior modifica-
tion field. In addition, scientific journals were developed to publish research in
behavior analysis and behavior modification, and professional organizations were
started to support research and professional activity in behavior analysis and behav-
ior modification. These books, journals, and organizations are listed in the timeline
in Figure 1-3. (For a more complete description of these publications and organi-
zations, see Cooper, Heron, and Heward [1987; 2007] and Michael [1993a].)

Introduction to Behavior Modification 9

Areas of Application

Behavior modification procedures have been used in many areas to help people
change a vast array of problematic behaviors (Carr & Austin, 2001; Gambrill,
1977; Lutzker & Martin, 1981; Vollmer, Borrero, Wright, Van Camp, & Lalli,
2001). This section briefly reviews these areas of application.

Developmental Disabilities
More behavior modification research has been conducted in the field of develop-
mental disabilities than perhaps any other area (Iwata et al., 1997). People with
developmental disabilities often have serious behavioral deficits, and behavior
modification has been used to teach a variety of functional skills to overcome
these deficits (Repp, 1983). In addition, people with developmental disabilities
may exhibit serious problem behaviors such as self-injurious behaviors, aggressive
behaviors, and destructive behaviors. A wealth of research in behavior modifica-
tion demonstrates that these behaviors often can be controlled or eliminated with
behavioral interventions (Barrett, 1986; Repp & Horner, 1999; Van Houten &

Skinner’s
basic research
on principles
of behavior

Skinner,
Walden Two

Skinner,
Science and
Human Behavior

SEAB

JEAB

Behavior
Research and
Therapy

Keller and
Schoenfeld,
Principles of
Psychology

Holland and
Skinner, The
Analysis of
Behavior

Ferster and
Skinner,
Schedules of
Reinforcement

Skinner,
Verbal Behavior

1958195319481930s 1963

1961195719501938 1966

AABTSkinner,
The Behavior
of Organisms

FIGURE 1-3 This timeline shows the major events in the development of behavior modification. Starting in
the 1930s with Skinner’s basic research on the principles of behavior, the timeline includes major
books, journals, and professional organizations. SEAB, Society for the Experimental Analysis of
Behavior; JEAB, Journal of the Experimental Analysis of Behavior; AABT, Association for Advance-
ment of Behavior Therapy; JABA, Journal of Applied Behavior Analysis.

10 Chapter 1

Axelrod, 1993; Whitman, Scibak, & Reid, 1983; Williams, 2004). Behavior modi-
fication procedures also are used widely in staff training and staff management in
the field of developmental disabilities (Reid, Parsons, & Green, 1989).

Mental Illness
Some of the earliest research in behavior modification demonstrated its effective-
ness in helping people with mental illness in institutional settings (Ayllon, 1963;
Ayllon & Michael, 1959). Behavior modification has been used with patients
with chronic mental illness to modify such behaviors as daily living skills, social
behavior, aggressive behavior, treatment compliance, psychotic behaviors, and
work skills (Dixon & Holcomb, 2000; Scotti, McMorrow, & Trawitzki, 1993;
Wilder, Masuda, O’Connor, & Baham, 2001). One particularly important contri-
bution of behavior modification was the development of a motivational procedure
for institutional patients called a token economy (Ayllon & Azrin, 1968). Token
economies are still widely used in a variety of treatment settings (Kazdin, 1982).

Education and Special Education
Behavior modification procedures are used widely in education (Alberto &
Troutman, 2003) and great strides have been made in the field of education
because of behavior modification research (Bijou & Ruiz, 1981). Researchers
have analyzed student–teacher interactions in the classroom, improved teaching
methods, and developed procedures for reducing problem behaviors in the class-
room (Bambara & Kern, 2003; Becker & Carnine, 1981; Madsen, Becker, &
Thomas, 1968; Sugai & Horner, 2005; Thomas, Becker, & Armstrong, 1968).

Skinner,
About
Behaviorism

Skinner,
Technology
of Teaching

Ayllon and
Azrin, The
Token Economy Continued

research and
publications
in behavior
modification

Skinner,
Contingencies
of Reinforcement:
A Theoretical
Analysis

Skinner,
Beyond Freedom
and Dignity

The Behavior
Analyst

Association for
Behavior Analysis

Progress in
Behavior
Modification

JABA
Behavior
Modification

Behavior
Therapy

Journal of
Organizational
Behavior
Management

Journal of
Behavior Therapy
and Experimental
Psychiatry

1974 19771968
1980s–
2000s

1978197519711969

1970

Emphasis on
Functional
Analysis

National
Certification
in Behavior
Analysis

Introduction to Behavior Modification 11

Behavior modification procedures have also been used in higher education to
improve instructional techniques, and thus improve student learning (Michael,
1991; Saville & Zinn, 2009).

In special education, that is, the education of people with developmental
disabilities or other special needs, behavior modification has played a major role
(Rusch et al., 1988) in developing teaching methods, controlling problem beha-
viors in the classroom, improving social behaviors and functional skills, promoting
self-management, and training teachers.

Rehabilitation
Rehabilitation is the process of helping people regain normal function after an
injury or trauma, such as a head injury from an accident or brain damage from a
stroke. Behavior modification is used in rehabilitation to promote compliance
with rehabilitation routines such as physical therapy, to teach new skills that can
replace skills lost through the injury or trauma, to decrease problem behaviors, to
help manage chronic pain, and to improve memory performance (Bakke et al.,
1994; Davis & Chittum, 1994; Heinicke, Carr, & Mozzoni, 2009; O’Neill &
Gardner, 1983; Tasky, Rudrud, Schulze, & Rapp, 2008).

Community Psychology
Within community psychology, behavioral interventions are designed to influence
the behavior of large numbers of people in ways that benefit everybody. Some tar-
gets of behavioral community interventions include reducing littering, increasing
recycling, reducing energy consumption, reducing unsafe driving, reducing illegal
drug use, increasing the use of seat belts, decreasing illegal parking in spaces for
the disabled, and reducing speeding (Cope & Allred, 1991; Cox & Geller, 2010;
Geller & Hahn, 1984; Ludwig & Geller, 1991; Van Houten & Nau, 1981; Van
Houten, Van Houten, & Malenfant, 2007).

Clinical Psychology
In clinical psychology, psychological principles and procedures are applied to help
people with personal problems. Typically, clinical psychology involves individual
or group therapy conducted by a psychologist. Behavior modification in clinical
psychology, often called behavior therapy, has been applied to the treatment of a
wide range of human problems (Hersen & Bellack, 1985; Hersen & Rosqvist,
2005; Hersen & Van Hasselt, 1987; Spiegler & Guevremont, 2010; Turner,
Calhoun, & Adams, 1981). Behavior modification procedures have also been
used to train clinical psychologists (Veltum & Miltenberger, 1989).

Business, Industry, and Human Services
The use of behavior modification in the field of business, industry, and human ser-
vices is called organizational behavior modification or organizational behavior man-
agement (Bailey & Burch, 2010; Daniels, 2000; Frederickson, 1982; Luthans &
Kreitner, 1985; Reid et al., 1989; Stajkovic & Luthans, 1997). Behavior modifica-
tion procedures have been used to improve work performance and job safety and
to decrease tardiness, absenteeism, and accidents on the job. In addition, behavior

12 Chapter 1

modification procedures have been used to improve supervisors’ performances. The
use of behavior modification in business and industry has resulted in increased pro-
ductivity and profits for organizations and increased job satisfaction for workers.

Self-Management
People use behavior modification procedures to manage their own behaviors.
They use self-management procedures to control personal habits, health-related
behaviors, professional behaviors, and personal problems (Brigham, 1989; Epstein,
1996; Stuart, 1977; Watson & Tharp, 1993; 2007; Yates, 1986). Chapter 20 dis-
cusses the application of behavior modification procedures for self-management.

Child Behavior Management
Numerous applications of behavior modification to the management of child behav-
ior exist (Durand & Hieneman, 2008; Hieneman, Childs, & Sergay, 2006; Miller,
1975; Patterson, 1975; Schaeffer & Millman, 1981). Parents and teachers can learn
to use behavior modification procedures to help children overcome bed wetting,
nail biting, temper tantrums, noncompliance, aggressive behaviors, bad manners,
stuttering, and other common problems (Christophersen & Mortweet, 2001; Gross &
Drabman, 2005; Watson & Gresham, 1998).

Prevention
Behavior modification procedures have been applied to preventing problems in
childhood (Roberts & Peterson, 1984). Other applications of behavior modification
in the area of prevention include preventing child sexual abuse, child abduction,
accidents in the home, child abuse and neglect, poisoning, infections, and sexually
transmitted diseases (Beck & Miltenberger, 2009; Carroll, Miltenberger, & O’Neill,
1992; Dancho, Thompson, & Rhoades, 2008; Montesinos, Frisch, Greene, &
Hamilton, 1990; Poche, Yoder, & Miltenberger, 1988). Preventing problems in the
community with behavior modification is one aspect of community psychology.

Sports Performance
Behavior modification is used widely in the field of sports psychology (Martin &
Hrycaiko, 1983). Behavior modification procedures have been used to improve ath-
letic performance in a wide variety of sports during practice and in competition
(Boyer, Miltenberger, Fogel, & Batsche, 2009; Brobst & Ward, 2002; Hume &
Crossman, 1992; Kendall, Hrycaiko, Martin, & Kendall, 1990; Wolko, Hrycaiko, &
Martin, 1993; Zeigler, 1994). Behavior modification procedures have been shown
to result in better athletic performance than do traditional coaching procedures.

Health-Related Behaviors
Behavior modification procedures are used to promote health-related behaviors by
increasing healthy lifestyle behaviors (such as exercise and proper nutrition) and decreas-
ing unhealthy behaviors (such as smoking, drinking, and overeating). Behavior modifi-
cation procedures are also used to promote behaviors that have a positive influence on
physical or medical problems—such as decreasing frequency and intensity
of headaches, lowering blood pressure, and reducing gastrointestinal disturbances
(Blumenthal & McKee, 1987; Dallery, Meredith, & Glenn, 2008; Gentry, 1984;

Introduction to Behavior Modification 13

Reynolds, Dallery, Shroff, Patak, & Lerass, 2008; Van Wormer, 2004)—and to increase
compliance with medical regimens (Levy, 1987). Applying behavior modification to
health-related behaviors is also called behavioral medicine or health psychology.

Gerontology
Behavior modification procedures are applied in nursing homes and other care
facilities to help manage the behavior of older adults (Hussian, 1981; Hussian &
Davis, 1985). Behavior modification procedures are used to help older adults deal
with their declining physical abilities, to help them adjust to nursing home envir-
onments, to promote health-related behaviors and appropriate social interactions,
and to decrease problem behaviors that may arise from Alzheimer’s disease, other
types of dementia, or institutional demands (Carstensen & Erickson, 1986; Dwyer-
Moore & Dixon, 2007; Moore, Delaney, & Dixon, 2007; Stock & Milan, 1993).

The Structure of This Textbook

This textbook is divided into five major sections. These sections discuss the follow-
ing topics:

â–  Measurement of behavior and behavior change
â–  Basic principles of behavior
â–  Procedures to establish new behaviors
â–  Procedures to decrease undesirable behaviors and increase desirable behaviors
â–  Other behavior change procedures

The book is designed so that the information presented in earlier sections is
applied in later sections.

Measurement of Behavior and Behavior Change
There are two chapters in Part 1 of this textbook. Chapter 2 teaches you how to
observe and record behaviors that are to be modified in a behavior modification
program. Chapter 3 teaches you how to construct graphs and evaluate graphed
data to analyze behavior change resulting from a behavior modification program.

Basic Principles of Behavior
The five chapters in Part 2 of this textbook discuss the basic principles of behavior
modification derived from scientific research in behavior analysis. The behavior
modification procedures discussed in the remainder of the book are based on the
basic behavioral principles reviewed in this section, which include reinforcement,
extinction, punishment, stimulus control, and respondent conditioning. Once you
understand these basic principles, it will be easier to understand and apply the
behavior modification procedures described in later sections.

Procedures to Establish New Behaviors
One goal of behavior modification is to establish desirable new behaviors or skills.
The four chapters in Part 3 of this textbook discuss behavior modification

14 Chapter 1

procedures used to establish new behaviors: shaping, prompting and transfer of
stimulus control, chaining, and behavioral skills training procedures.

Procedures to Increase Desirable Behaviors and Decrease
Undesirable Behaviors
Another goal of behavior modification procedures is to decrease the occurrence of
undesirable behaviors and increase the occurrence of desirable behaviors that are
not occurring frequently enough. The occurrence of undesirable behaviors is a
behavioral excess. Desirable behaviors that occur too infrequently are behavioral
deficits. The seven chapters in Part 4 describe how to analyze behaviors and how
to apply reinforcement, extinction, stimulus control, and punishment to decrease
excess behaviors while increasing more desirable behaviors.

Other Behavior Change Procedures
The six chapters in Part 5 of this textbook describe more complex behavior
modification procedures. Chapter 20 discusses self-management procedures.
Chapter 21 discusses habit disorders and procedures for decreasing these excess
behaviors. Chapter 22 on token economies and Chapter 23 on behavioral con-
tracting discuss procedures that extend the reinforcement and punishment proce-
dures described earlier. Chapter 24 applies procedures based on respondent
conditioning to decrease fear and anxiety. Chapter 25 discusses behavior modifica-
tion procedures to change cognitive behaviors, a type of covert behavior.

CHAPTER SUMMARY

1. Human behavior is defined as an individual’s
actions that have one or more physical dimensions
and can be observed and recorded. Behaviors have
an impact on the physical or social environment.
Behavior is lawful; its occurrence is influenced by
environmental events. A behavior may be overt or
covert.

2. Behavior modification procedures involve analyz-
ing and manipulating current environmental
events to change behavior. A behavioral excess
or behavioral deficit may be targeted for change
with behavior modification procedures. Behavior
modification procedures are based on behavioral
principles derived from scientific research. B. F.
Skinner conducted the early scientific research
that laid the foundation for behavior modification.
He also published a number of books demonstrating
the application of behavioral principles to everyday
life. Behavior modification procedures often are
implemented by people in everyday life. Behavior

is measured before and after the behavior
modification procedures are applied to document
the effectiveness of the procedures. Behavior modifi-
cation de-emphasizes past events and rejects hypo-
thetical underlying causes of behavior.

3. The historical roots of behavior modification
can be found in the work of Pavlov, Thorndike,
Watson, and especially B. F. Skinner, who identi-
fied a number of basic principles of behavior and
wrote about applying the principles of behavior
analysis to human behavior.

4. Behavior modification procedures have been
applied successfully to all aspects of human
behavior, including developmental disabilities;
mental illness; education and special education;
rehabilitation; community psychology; clinical psy-
chology; business, industry, and human services;
self-management; child behavior management;
prevention; sports performance; health-related
behaviors; and gerontology.

Introduction to Behavior Modification 15

KEY TERMS

applied behavior analysis, 5
behavior, 2
behavior modification, 5
behavioral deficit, 5
behavioral excess, 5
controlling variable, 5

covert behaviors, 3
dimensions, 2
duration, 2
experimental analysis of

behavior, 5
frequency, 2

intensity, 2
latency, 2
law of effect, 9
overt behavior, 3
target behavior, 5

PRACTICE TEST

1. What is behavior? (p. 2)
2. Provide an example of a description of behavior

and the label applied to that behavior. (p. 2)
3. Describe the four physical dimensions of behav-

ior that can be observed and recorded. (p. 2)
4. Provide an example of how a behavior has an

impact on the physical environment. Provide
an example of how a behavior has an impact
on the social environment. (p. 2)

5. What does it mean to say that behavior is lawful?
What is a functional relationship? (p. 2)

6. Describe the distinction between overt behavior
and covert behavior. Provide an example of
each. Which type of behavior is the focus of
this book? (p. 3)

7. Identify the six characteristics of human behav-
ior. (pp. 2–3)

8. What does it mean to say that behavior modifi-
cation procedures are based on behavioral
principles? (p. 5)

9. What causes human behavior? Describe how a
label might be mistakenly identified as a cause
of a behavior. (pp. 5–6)

10. Why is it important to describe behavior modifi-
cation procedures precisely? (p. 6)

11. Who implements behavior modification proce-
dures? (p. 6)

12. Why is it important to measure behavior before
and after behavior modification procedures are
used? (pp. 6–7)

13. Why doesn’t behavior modification focus on the
past as the cause of the behavior? (p. 7)

14. Identify eight defining characteristics of behav-
ior modification. (pp. 5–7)

15. Briefly describe the contributions of Pavlov,
Thorndike, Watson, and Skinner to the develop-
ment of behavior modification. (pp. 8–9)

16. Identify at least one way in which behavior
modification has been applied in each of the fol-
lowing areas: developmental disabilities; educa-
tion; community psychology; business, industry,
and human services; self-management; preven-
tion; health-related behaviors; mental illness; reha-
bilitation; clinical psychology; child management;
sports performance; and gerontology. (pp. 10–14)

16 Chapter 1

Chapter Two

Observing and Recording Behavior

One fundamental aspect of behavior modification is measuring the behaviorthat is targeted for change. Measurement of the target behavior (or beha-
viors) in behavior modification is called behavioral assessment. Behavioral assess-
ment is important for a number of reasons.

â–  Measuring the behavior before treatment provides information that can
help you determine whether treatment is necessary.

â–  Behavioral assessment can provide information that helps you choose the
best treatment.

â–  Measuring the target behavior before and after treatment allows you to
determine whether the behavior changed after the treatment was
implemented.

Consider the following example.
A supervisor in a manufacturing plant believed the company

had a problem with workers showing up late for work. Before taking
any remedial action, the supervisor recorded the arrival times of the
workers for a number of days (Figure 2-1). The assessment showed
that there were few instances of tardiness. In this case, behavioral
assessment demonstrated that there was not a problem and that
intervention was not necessary.

If the measurement of the workers’ arrival times showed that
there was a problem, the supervisor would develop a behavior modifi-
cation plan to change the workers’ behavior. The supervisor would
continue to record arrival times as the intervention was implemented.
The measurement of the workers’ arrival times before, during, and
after intervention would demonstrate whether the workers arrived
late less frequently once intervention had been implemented.

Direct and Indirect Assessment

There are two types of behavioral assessment: direct and indirect (Iwata, Vollmer, &
Zarcone, 1990; Martin & Pear, 1999; O’Neill et al., 1997). Indirect assessment

â–  How do you define a target behavior in
a behavior modification program?

â–  What different methods can you use to
record a target behavior?

â–  How does continuous recording differ
from interval and time sample
recording?

â–  What is reactivity of behavior recording,
and how can you minimize it?

â–  What is interobserver reliability, and
why is it important?

17

involves using interviews, questionnaires, and rating scales to obtain information on
the target behavior from the person exhibiting the behavior or from others (e.g., par-
ents, teachers, or staff). Indirect assessment does not occur when the target behavior
occurs but relies on an individual’s recall of the target behavior. With direct assess-
ment, a person observes and records the target behavior as it occurs. To observe the
target behavior, the observer (or a video camera, in some cases) must be in close
proximity to the person exhibiting the behavior so that the target behavior can be
seen (or heard). In addition, the observer must have a precise definition of the target
behavior so that its occurrence can be distinguished from occurrences of other
behaviors. To record the target behavior, the observer must register the occurrence
of the behavior when it is observed; various methods of recording are described
later in this chapter. When a school psychologist observes a socially withdrawn
child on the playground and records each social interaction with another child, the
psychologist is using direct assessment. When the psychologist interviews the stu-
dent’s teacher and asks the teacher how many times the child usually interacts with
other children on the playground, the psychologist is using indirect assessment.

Direct assessment usually is more accurate than indirect assessment. This is
because in direct assessment, the observer is trained specifically to observe the tar-
get behavior and record its occurrence immediately. In indirect assessment, infor-
mation on the target behavior depends on people’s memories. In addition, the
people providing information may not have been trained to observe the target
behavior and may not have noticed all the occurrences of the behavior. As a
result, indirect assessment may be based on incomplete information about the tar-
get behavior. Therefore, most research and application in behavior modification
relies on direct assessment.

The remainder of this chapter discusses direct assessment methods for
observing and recording the target behavior in a behavior modification program,

FIGURE 2-1 The supervisor collects data on the number of workers who arrive late.

18 Chapter 2

specifically the steps needed to develop a behavior recording plan. These steps
include the following:

1. Defining the target behavior
2. Determining the logistics of recording
3. Choosing a recording method
4. Choosing a recording instrument

Defining the Target Behavior

The first step in developing a behavior recording plan is to define the target behavior
you want to record. To define the target behavior for a particular person, you must
identify exactly what the person says or does that constitutes the behavioral excess or
deficit targeted for change. A behavioral definition includes active verbs describing spe-
cific behaviors that a person exhibits. A behavioral definition is objective and unambig-
uous. As an example of defining a target behavior, unsportsmanlike behavior for a
particular baseball player may be defined as yelling obscenities, throwing the bat or bat-
ting helmet, and kicking the dirt as the player walks back to the bench after striking out.

Note that the example does not refer to any internal states such as being
angry, upset, or sad. Such internal states cannot be observed and recorded by
another person. The behavioral definition does not make inferences about a per-
son’s intentions. Intentions cannot be observed, and inferences about intentions
often are incorrect. Finally, a label (“a bad sport”) is not used to define the behav-
ior because labels do not identify the person’s actions.

Labels for behaviors are ambiguous; they can mean different things to different
people. For example, to one person, unsportsmanlike behavior might mean fighting
with a member of the other team, whereas another person considers it to mean
cursing, throwing a bat, and kicking dirt. Specific behaviors can be observed and
recorded; labels for the behavior cannot. In addition, labels can be used incorrectly
as explanations of a behavior. For example, if a person is observed to repeat syllables
or words when he talks, we might label him a stutterer. To then say that the person
repeats syllables or words because he is a stutterer is an incorrect use of the label as
a cause of the behavior. Repeating words or syllables is not caused by stuttering; it is
a behavior called stuttering. The main value of labels is that they may be used as
convenient shorthand when referring to a target behavior. However, the behavior
must always be defined before it can be observed and recorded.

One characteristic of a good behavioral definition is that after seeing the defi-
nition, different people can observe the same behavior and agree that the behavior
is occurring. When two people independently observe the same behavior and
both record that the behavior occurred, this is called interobserver agreement
(IOA) or interobserver reliability (Bailey, 1977; Bailey & Burch, 2002). IOA,
which is commonly reported in behavior modification research, is discussed in
more detail later in this chapter.

Table 2-1 lists behavioral definitions for common target behaviors and the
labels associated with those behaviors. The behaviors that are described could be
observed and agreed on by two independent observers. The labels, in contrast,
are general names that are commonly used for these types of behaviors. Labels

Observing and Recording Behavior 19

such as these may also be used to refer to behaviors other than those defined here.
For example, in contrast with the definition given for Bobby in Table 2-1, a tan-
trum could be a label for the behavior of screaming, cursing at parents, slamming
doors, and throwing toys on the floor. You must develop a specific behavioral def-
inition that fits the target behavior of the person you are observing.

Researchers in behavior modification carefully define the target behaviors of
people for whom they provide treatment. For example, Iwata and his colleagues
(Iwata, Pace, Kalsher, Cowdery, & Cataldo, 1990) used behavior modification pro-
cedures to decrease self-injurious behavior in children with intellectual disabilities.
Their definitions for three types of self-injurious behavior were as follows: “arm bit-
ing—closure of upper and lower teeth on any portion of the skin extending from
fingers to elbow; face hitting—audible contact of an open or closed hand against
the face or head; and head banging—audible contact of any portion of the head
against a stationary object (e.g., desk, floor, wall)” (p. 13). In another example,
Rogers-Warren, Warren, and Baer (1977) used behavior modification procedures to
increase sharing in preschool children. They defined sharing as occurring “when
one subject passed or handed a material to a second subject, when subjects
exchanged materials, or when two or more subjects simultaneously used the same
material (for example, two subjects coloring on the same piece of paper)” (p. 311).

FOR FURTHER READING
Social Validity

When using behavior modification procedures to help people change their behavior, it is important to
choose target behaviors that are socially significant; behaviors that the client agrees are important tar-
gets for change. One way to make sure you are choosing important (socially significant) target behaviors
is to assess the opinions of the client or other important individuals (parents, teachers, etc.). When such
individuals agree that the target behaviors are important and acceptable, they are establishing the social
validity of the target behaviors. Kazdin (1977) and Wolf (1978) discussed the importance of social validity
in behavior modification and methods for assessing social validity.

TABLE 2-1 Behavioral Definitions and Labels for Common Problems

Behavioral Definition Label

When Bobby cries and sobs, lies on the floor and kicks the floor or walls, or pounds
toys or other objects on the floor, it is defined as a tantrum.

Tantrumming

Studying for Rae involves reading pages from a textbook, underlining sentences in the
text, completing math or physics workbook exercises, reading notes from class, and
outlining chapters from the text.

Studying

When Pat says no to someone who asks her to do something that is not part of her job,
when she asks coworkers not to smoke in her office, and when she asks coworkers to
knock before entering her office, it is defined as assertiveness.

Assertiveness

Stuttering is defined for Joel as repeating a word or a word sound, prolonging the sound
when saying a word, or hesitating more than 2 seconds between words in a sentence or
between syllables in a word.

Stuttering

Any time Mark’s finger is in his mouth and his teeth are closed together on the finger-
nail, cuticle, or skin around the nail, it is defined as nail-biting.

Nail-biting

20 Chapter 2

The Logistics of Recording

The Observer
We have defined the target behavior to be recorded for a client, that is, a person
who exhibits the target behavior and with whom the behavior modification pro-
gram will be implemented. The next step is to identify who will observe and
record the behavior. In a behavior modification program, the target behavior typi-
cally is observed and recorded by a person other than the one exhibiting the target
behavior (i.e., an independent observer). The observer may be a professional, such
as a behavior analyst or a psychologist, or a person routinely associated with the
client in the client’s natural environment, such as a teacher, parent, staff member,
or supervisor. The observer must have proximity to the client to observe the target
behavior when it occurs. The exception would be when the target behavior is
observed via video. The observer must be trained to identify the occurrence of
the target behavior and to record the behavior immediately. He or she also must
have the time to observe and record the behavior and must be willing to function
as an observer. For example, a teacher may be asked to observe and record the tar-
get behavior of one of her students, but may not agree to do so because the
demands of teaching her students do not allow her the time to function as an
observer. In most cases, it is possible to develop a behavior recording plan such
that a person can observe and record the target behavior of the client without too
much disruption of his or her normal routine.

In some cases, the observer is the person exhibiting the target behavior. When
the client observes and records his or her own target behavior, it is called self-
monitoring. Self-monitoring is valuable when it is not possible for another
observer to record the target behavior, as when the target behavior occurs infre-
quently or when it occurs only when no one else is present (Stickney & Milten-
berger, 1999; Stickney, Miltenberger, & Wolff, 1999). Self-monitoring may also
be combined with direct observation by another observer. For example, a psychol-
ogist might directly observe and record the behavior of a person who is receiving
treatment for a nervous habit such as hair-pulling. In addition, the client might
be asked to self-monitor the target behavior outside the therapy sessions. If self-
monitoring is used in a behavior modification program, the client must be trained
to record his or her own behavior in the same way that an observer would be
trained.

When and Where to Record
The observer records the target behavior in a specific period called the observa-
tion period. It is important to choose an observation period at the time when the
target behavior is likely to occur. Indirect assessment information from the client
or others (e.g., from an interview) may indicate the best times to schedule the
observation period. For example, if staff members report that a patient in a psychi-
atric ward is most likely to engage in disruptive behavior (defined as screaming,
pacing, and cursing at other residents) around mealtimes, the observation period
would be scheduled during meals. The timing of the observation periods also is
determined by the availability of the observer(s) and the constraints imposed by

Observing and Recording Behavior 21

the client’s activities or preferences. Note that the client or the client’s parent or
guardian must give consent before you can observe and record his or her behav-
ior. This is particularly important when observation occurs without the client’s
knowledge. In such cases, the client must provide consent for observations to
occur, with the understanding that some observations may occur at times
unknown to him or her (e. g., Wright & Miltenberger, 1987).

Observation and recording of behavior take place in natural settings or in ana-
logue settings. A natural setting consists of the places in which the target behavior
typically occurs. Observing and recording a target behavior in the classroom is an
example of a natural setting for a student. Observing a target behavior in a clinic
playroom is an analogue setting because being in the clinic is not part of the
child’s normal daily routine. Observation in a natural setting is likely to provide a
more representative sample of the target behavior. The target behavior may be
influenced by the analogue setting, and observation in this setting may provide a
sample that is not representative of the behavior under normal circumstances.
However, there are benefits of observing in an analogue setting: It is more con-
trolled than a natural setting, and the variables that influence the behavior are eas-
ier to manipulate.

Observation of the target behavior can be structured or unstructured. When
observations are structured, the observer arranges for specific events or activities to
occur during the observation period. For example, when observing child behavior
problems, the observer may ask the parent to make specific requests of the child
during the observation period. During unstructured observations, no specific
events or activities are arranged and no instructions are given during the observa-
tion period.

When self-monitoring is used, the client may be able to observe and record
the target behavior throughout the day and may not be constrained by a specific
observation period. For example, clients who are self-monitoring the number of
cigarettes they smoke each day can record each cigarette smoked regardless of
when they smoke it. However, some behaviors may occur with such frequency
that the client could not record continuously throughout the day; for example, a
client who stutters may engage in stuttering hundreds of times throughout the
day. In cases such as this, the client would be instructed to record the behavior
during observation periods agreed on in advance with the psychologist.

In behavior modification research, the people observing and recording the tar-
get behaviors usually are trained research assistants. They study the behavioral def-
inition of the target behavior and then practice recording under the supervision of
the researcher. When they can record the behavior reliably during practice ses-
sions (after they have good IOA with the researcher), they record the target behav-
ior during actual observation periods as part of the study. The observation periods
used in behavior modification research often are brief (say, 15-30 minutes). When
observations occur in natural settings, researchers usually choose observation peri-
ods that are representative of the usual occurrence of the target behavior. For
example, observations may take place in a classroom, workplace, hospital, or
other setting in which the target behavior usually occurs. In a study using behavior
modification to improve children’s behavior during trips to the dentist, Allen and
Stokes (1987) recorded children’s disruptive behavior (defined as head and body

22 Chapter 2

movements, crying, gagging, and moaning) while they were in the dentist’s chair
and the dentist performed dental procedures on them. In another study, Durand
and Mindell (1990) taught parents how to use behavior modification procedures
to decrease nighttime tantrum behavior (defined as loud screaming and hitting
furniture) in their young child. In this study, the parents recorded the target beha-
viors for an hour before the child’s bedtime because this was the time when the
tantrum behaviors occurred.

When observations occur in analogue settings, researchers often simulate
events that are likely to occur in natural settings. For example, Iwata, Dorsey,
Slifer, Bauman, and Richman (1982) observed and recorded the self-injurious behav-
ior of children with intellectual disabilities in therapy rooms in a hospital. During
their observation periods, they simulated different events or activities that the chil-
dren were likely to experience at home or at school. For example, the researchers
observed the children as they played with toys, as teachers gave them instructions,
and during times they were receiving no attention from the teacher. Iwata and his
colleagues found that for each child, the self-injurious behavior occurred at different
rates in observation periods that simulated different events or activities.

Choosing a Recording Method

Different aspects of the target behavior may be measured using different recording
methods. These methods include continuous recording, product recording, inter-
val recording, and time sample recording. Each method is described here.

Continuous Recording
In continuous recording, the observer observes the client continuously through-
out the observation period and records each occurrence of the behavior. To do
so, the observer must be able to identify the onset and the offset (or beginning
and end) of each instance of the behavior. In continuous recording, the observer
can record various dimensions of the target behavior, particularly its frequency,
duration, intensity, and latency.

The frequency of a behavior is the number of times the behavior occurs in
an observation period. You measure the frequency of a behavior simply by count-
ing each time that it occurs. One occurrence is defined as one onset and offset of
the behavior. For example, you can count the number of cigarettes someone
smokes. For this target behavior, the onset may be defined as lighting the cigarette
and the offset as putting it out. You will use a frequency measure when the num-
ber of times the behavior occurs is the most important information about the
behavior. Frequency may be reported as rate, which is frequency divided by the
time of the observation period. Rate is often reported as responses per minute.

The duration of a behavior is the total amount of time occupied by the
behavior from start to finish. You measure the duration of a behavior by timing it
from its onset to its offset. For example, you might record the number of minutes
a student studies per day, the number of minutes a person exercises, or the num-
ber of seconds a patient who has had a stroke stands up without assistance during
rehabilitation sessions in the hospital. You will use a duration measure when the

Observing and Recording Behavior 23

most important aspect of the behavior is how long it lasts. Duration may be
reported as percentage of time, which is duration divided by the time of the obser-
vation period (Miltenberger, Rapp, & Long, 1999).

Some researchers use a real-time recording method in which the exact time
of each onset and offset of the target behavior is recorded (Miltenberger et al.,
1999; Miltenberger, Long, Rapp, Lumley, & Elliott, 1998). With real-time record-
ing, the researchers have a record of the frequency and duration of the target
behavior, as well as the exact timing of each occurrence of the behavior. Real-
time recording can be carried out after video recording the target behavior in the
observation period. The observer then plays the video and records the time indi-
cated on the timer at the onset and offset of each occurrence of the behavior on
a data sheet developed for real-time recording (Rapp, Carr, Miltenberger, Dozier,
& Kellum, 2001). Alternatively, handheld or laptop computers with software that
permits recording of the exact timing of events can be used for real-time recording
(Kahng & Iwata, 1998).

The intensity of a behavior is the amount of force, energy, or exertion
involved in it. Intensity (also called magnitude) is more difficult to measure than
frequency or duration because it does not involve simply counting the number of
times the behavior occurs or recording the amount of time it takes to occur. Inten-
sity often is recorded with a measurement instrument or by using a rating scale.
For example, you could use a decibel meter to measure the loudness of someone’s
speech. A physical therapist might measure the strength of a person’s grip to judge
recovery from an injury. Parents might use a rating scale from 1 to 5 to measure
the intensity of a child’s tantrum. The parents would have to define the behavior
associated with each point on the rating scale so that their ratings were reliable;
their ratings would be reliable if they both observed a tantrum and recorded the
same number on the rating scale. Intensity is not used as often as frequency or
duration, but it is a useful measure when you are most interested in the force or
magnitude of the behavior (Bailey, 1977; Bailey & Burch, 2002).

The latency of the behavior is the time from some stimulus or event to the
onset of the behavior. You measure latency by recording how long it takes the per-
son to initiate the behavior after a particular event occurs. For example, you could
record how long it takes a child to start putting toys away after being asked to do
so. The shorter the latency, the sooner the child initiates the behavior after the
request. Another example of latency is the time it takes a person to answer the
phone after it starts ringing.

How does latency differ from duration?

Latency is the time from some stimulus or event to the onset of the behavior,
whereas duration is the time from the onset of the behavior to its offset. That is,
latency is how long it takes to start the behavior and duration is how long the
behavior lasts.

When using continuous recording, you can choose one or more dimensions
to measure. The dimension you choose depends on which aspect of the behavior
is most important and which dimension is most sensitive to change in the behav-
ior after treatment. For example, if you want to record a person’s stuttering, fre-
quency may be the most important dimension because you are interested in the

24 Chapter 2

number of stuttered words. You can then compare the number of stuttered words
before, during, and after treatment. If treatment is successful, there should be
fewer stuttered words. However, duration may also be an important dimension of
stuttering if there are long speech blocks or prolongations. In this case, you
would expect the duration of stuttering to decrease after treatment.

If you were recording a child’s tantrum behavior (screaming, throwing toys,
slamming doors), which dimension of the behavior would you measure?

The example of a child’s tantrum behavior is less clear. You may be interested
in the number of tantrums per day (frequency), but you may also be interested in
how long each tantrum lasts (duration). Finally, you may be interested in how
loud the child screams or how forcefully the child throws toys or slams doors
(intensity). We hope that, after treatment, the tantrums will decrease in frequency,
duration, and intensity; that is, they will occur less often, will not last as long, and
will not be as loud or violent.

Unless you measure the right dimension of a behavior, you may not be able
to judge the effectiveness of treatment. If you are in doubt, or if multiple dimen-
sions of the behavior seem relevant, the best course of action is to measure more
than one dimension. Go back to the example of the child’s tantrums. Figure 2-2
shows that, from an average of more than six per day during baseline, the fre-
quency of tantrums decreased to less than two per day during treatment. (Base-
line is the period during which the target behavior is recorded before treatment
is implemented.) It appears that treatment was effective. However, Figure 2-3
shows the duration of tantrums before and during treatment. Before treatment,
each of the five to eight tantrums per day lasted about 1 minute each, for a
total of 5 to 8 minutes of tantrum behavior per day. During treatment, the dura-
tion of each tantrum was much longer, resulting in more minutes of tantrum

FIGURE 2-2 The frequency of tantrums during baseline and treatment phases. During the baseline phase, the
target behavior is recorded, but treatment is not yet implemented. Tantrums decreased from an
average of more than six per day during baseline to less than two per day during treatment.

Observing and Recording Behavior 25

behavior per day. Therefore, according to the duration measure, the tantrums
got worse during treatment. This underscores the importance of measuring
more than one dimension of a target behavior because more than one dimen-
sion can change after treatment.

Note also that, to demonstrate the effectiveness of treatment, you must use
established research methods and an experimental design. Simply measuring the
behavior before, during, and after treatment demonstrates whether the target
behavior changed, but does not prove that the treatment caused the behavior
change (see Chapter 3).

Percentage of Opportunities
Percentage of trials or percentage correct is one final way in which event record-
ing may be conducted. In this method, the observer records the occurrence of a
behavior in relation to some other event, such as a learning trial or a response
opportunity, and reports the results as the percentage of opportunities in which
the behavior occurred. To say that a student complied with a teacher’s requests
11 times during the observation period or got 13 words correct on a spelling test
is inadequate information because there is no mention of response opportunities.
Reporting the results as the number of times the behavior occurred divided by
the number of opportunities provides more useful information. If the teacher
made 12 requests and the student complied with the teacher’s requests 11 times,
the percentage of compliance is 11/12, or 92%. However, if the teachers made
25 requests and the student complied 11 times, the percentage is only 44%, a
much less acceptable level of the behavior.

FIGURE 2-3 The duration of tantrums during baseline and treatment phases. Tantrums increased from an aver-
age duration of 1 minute each or a total of 5 to 8 minutes per day during baseline to about 6 min-
utes each or a total of 6 to 18 minutes per day during treatment. Therefore, the duration of tantrum
behavior per day did not decrease even though the frequency of tantrums decreased.

26 Chapter 2

Product Recording
Another aspect of a behavior that may be recorded is its product. Product record-
ing, also called permanent product recording (Marholin & Steinman, 1977), is an
indirect assessment method that can be used when a behavior results in a certain
tangible outcome that you are interested in. For example, a supervisor could
count the number of units assembled in a factory as a product measure of a work-
er’s job performance, or a teacher could record the number of correctly com-
pleted homework problems or workbook pages as a product measure of students’
academic performance (Noell et al., 2000). In their research on student behavior
problems and academic performance, Marholin and Steinman (1977) looked at
the math worksheets of students and recorded the number of math problems com-
pleted correctly as permanent products of the students’ academic performance.

One benefit of product recording is that the observer does not have to be
present when the behavior occurs. The teacher probably will not be present
when students complete their homework assignments, but he or she can still mea-
sure the product of the behavior (completed homework problems). One drawback
of product recording is that you cannot always determine who engaged in the
behavior that led to the product you recorded. For example, the teacher cannot
determine whether the students completed their own homework, whether some-
one else helped them, or whether someone did it for them.

Interval Recording
Another method for recording behavior is to record whether the behavior
occurred during consecutive time periods. This is called interval recording. To
use interval recording, the observer divides the observation period into a number
of smaller time periods or intervals, observes the client throughout each consecu-
tive interval, and then records whether the behavior occurred in that interval.
There are two types of interval recording: partial interval recording and whole
interval recording. With partial interval recording, you are not interested in the
number of times the behavior occurs (frequency) or how long it lasts (duration).
You do not have to identify the onset and offset of the behavior; rather, you simply
record whether the behavior occurred during each interval of time.

Suppose that a teacher is recording whether a child disrupts the class during
each 15-minute interval in the class period. The teacher sets a timer to beep every
15 minutes. When the disruptive behavior occurs, the teacher marks the corre-
sponding interval on a data sheet. Once an interval is marked, the teacher does not
have to observe the child or record the behavior until the next interval begins. If the
behavior does not occur in an interval, the interval is left blank on the data sheet.
Thus, one benefit of partial interval recording is that it takes less time and effort:
The observer records the behavior only once during the interval, regardless of how
many times the behavior occurs or how long it lasts. With whole interval recording,
the occurrence of the behavior is marked in an interval only when the behavior
occurs throughout the entire interval. If the behavior occurs in only part of the
interval, the behavior is not scored as occurring in that interval.

When researchers use interval recording, they often choose short intervals,
such as 6 or 10 seconds (Bailey, 1977; Bailey & Burch, 2002). In this way, they

Observing and Recording Behavior 27

make many recordings of the behavior during the observation period and obtain a
more representative sample of the target behavior than they could derive from lon-
ger intervals. For example, Iwata, Pace, Kalsher, Cowdery, and Cataldo (1990)
used 10-second intervals to record the occurrence of self-injurious behavior (e.g.,
head-banging, slapping, and scratching) in children with intellectual disabilities.
Miltenberger, Fuqua, and McKinley (1985) used 6-second intervals to record the
occurrence of motor tics (e.g., jerking movements of the head or facial muscles,
rapid eye-blinking) in adults. In this study, the researchers video recorded the
adults in the observation sessions and then recorded the number of intervals con-
taining motor tics from the videos. Every 6 seconds, the researchers recorded the
presence or absence of the tic behavior.

In some cases, frequency recording and interval recording can be combined
to produce frequency-within-interval recording. With this method, the observer
records the frequency of the target behavior but does so within consecutive inter-
vals of time in the observation period (Bailey, 1977; Bailey & Burch, 2002).
Frequency-within-interval recording shows you the frequency of the behavior and
the specific intervals in which the behavior occurred.

Time Sample Recording
When using time sample recording, you divide the observation period into inter-
vals of time, but you observe and record the behavior during only part of each inter-
val. The observation periods are separated by periods without observation. For
example, you might record the behavior for only 1 minute during each 15-minute
interval, or you might record the behavior only if it is occurring at the end of the
interval. Consider an observer who is using time sample recording to record a cli-
ent’s poor posture (defined as slouching, bending the back forward). The observer
sets a timer to beep every 10 minutes and records an instance of bad posture only
if the client’s posture is bad when the timer beeps at the end of the interval. Time
sample recording is valuable because the person does not have to observe the
behavior for the entire interval. Rather, the observer records the behavior that occurs
during only a portion of the interval or at a specific time in the interval.

In interval recording or time sample recording, the level of the behavior is
reported as the percentage of intervals in which the behavior occurred. To calcu-
late the percentage of intervals, you divide the number of scored intervals by the
total number of intervals during the observation period. A scored interval is an
interval in which the behavior was recorded.

Recording Methods
Continuous recording Record every instance of the behavior occurring during the observation period.

May record frequency, duration, intensity, or latency

Product recording Record the tangible outcome or permanent product of the occurrence of the
behavior

Interval recording Record the occurrence or nonoccurrence of the behavior in consecutive inter-
vals of time during an observation period

Time sample
recording

Record the occurrence or nonoccurrence of the behavior in discontinuous
intervals of time (time samples) during an observation period

28 Chapter 2

Choosing a Recording Instrument

The final step in developing a behavior recording plan is to choose a recording
instrument. The recording instrument is what the observer uses to register or
make a permanent product of the occurrence of the behavior. Paper and pencil
are used most often to record behavior. Put simply, the observer makes a note on
the paper each time he or she observes the behavior. To record behavior most
effectively, the observer uses a data sheet prepared in advance for the particular
behavior. The data sheet helps organize the recording process by making it clear
what the observer is to write down when the behavior occurs.

The data sheet in Figure 2-4 is used to record the frequency of a target behav-
ior. Each time the behavior occurs on a particular day, the observer marks an X in
one of the boxes for that day. The number of boxes with Xs filled in for each day
signifies the frequency, or the number of times that the behavior occurred on
each day.

The data sheet in Figure 2-5 is used to record the duration of a target behav-
ior. On each day, there are places to record the times the behavior started (onset)
and ended (offset). By recording the onset and offset of each instance of a behav-
ior, you end up with a recording of how long the behavior occurred (duration), as
well as how often it occurred (frequency).

An example of a data sheet used for 10-second interval recording is shown in
Figure 2-6. Notice that there are 6 boxes on each line and 15 lines of boxes. Each
box represents one 10-second interval, for a total of 90 intervals in 15 minutes. To
use the 10-second interval recording method, the observer listens to a recording
that signals the start of each interval. When the target behavior occurs, the

Name:
Observer:
Definition of behavior being recorded:

Date Frequency

Frequency Data Sheet

1 2 3 4 5 6 7 8 9 10 11 12

Daily Total

FIGURE 2-4 This data sheet is used to record the frequency of a behavior. You put an X into a box each time
the behavior occurs. If more than 12 instances of the behavior occur per day, continue recording
on the next line.

Observing and Recording Behavior 29

observer puts a check mark in the corresponding interval box. If the target behav-
ior does not occur during an interval, the observer leaves that interval box blank.
Alternatively, each interval box could have one or more codes. The observer cir-
cles or puts a check mark through the code that represents the behavior observed
in that interval. For example, the codes AT and RP could be used to signify the
behaviors of attention and reprimand, respectively, when observing a parent’s
behavior while interacting with a child. If the parent pays attention to the child
or reprimands the child in an interval, the observer would circle AT or RP, respec-
tively, for that interval.

Other procedures for recording behavior involve writing the behavior down
each time it occurs. For example, a person who wants to count the number of
cigarettes she smokes each day may keep a note card tucked into the cellophane
wrapper on the cigarette pack. Each time she smokes a cigarette, she makes a
check mark on the note card and counts the check marks at the end of each day.
Likewise, a person who is recording his rude behavior might keep a small note
pad in his shirt pocket; every time he makes a rude remark, he pulls out the note
pad and makes a note of it.

Not all instruments for recording behavior depend on paper and pencil. Any-
thing you can use to register each occurrence of a behavior can be considered a
behavior recording instrument. The following are some common examples.

â–  Use a golf stroke counter to record the frequency of a behavior. The
golf stroke counter is worn on the wrist like a wristwatch. Each time the
behavior occurs, you push the button on the counter (Lindsley, 1968).

â–  Use a stopwatch to record the cumulative duration of a behavior. You start
and stop the stopwatch each time the behavior starts and stops. Runners

Name:
Observer:
Definition of behavior being recorded:

Date

Onset Offset Onset Offset Onset Offset

Duration

Duration Data Sheet

Daily Duration

FIGURE 2-5 This data sheet is used to record the duration of a behavior. You record the onset and offset time
for each instance of the behavior. If there are more than three instances of the behavior per day,
continue recording on the next line.

30 Chapter 2

and joggers often wear watches with a stopwatch function that allows them
to record the duration of their workouts.

â–  Use a laptop or handheld computer, Palm Pilot or other PDA, or other
handheld electronic devices with an app for data recording to record the fre-
quency and duration of many behaviors at once. You push different keys on
the computer or handheld device each time different behaviors occur; as
long as you keep pressing the key, the duration of the behavior is recorded
(Dixon, 2003; Fogel, Miltenberger, Graves, & Koehler, in press; Iwata, Pace,
Kalsher, Cowdery, & Cataldo, 1990; Jackson & Dixon, 2007; Kahng &
Iwata, 1998).

â–  Use barcode technology for recording behavior. Each behavior being recorded
is given a unique barcode and the observer has a sheet of paper with all the
barcodes for the behaviors to be recorded. When a particular behavior occurs,
the observer scans the barcode for that behavior to record its occurrence.

Name:
Observer:
Date and time of observation:
Definition of behavior being recorded:

Minutes of observation

1 2 3 4 5 6

Ten-second intervals

Interval Data Sheet

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15

FIGURE 2-6 This is an interval recording data sheet. Each box corresponds to an interval, and a check mark is
placed in a box when the behavior occurs during that interval. When the behavior does not occur
during an interval, the box is left blank.

Observing and Recording Behavior 31

â–  Transfer a coin from one pocket to another to record the frequency of a
behavior. Each time you observe the behavior, you move a coin from your
right pocket to your left pocket. The number of coins in your left pocket at
the end of the day equals the frequency of the behavior (assuming that you
don’t spend any of the coins from your left pocket).

â–  Make small tears in a piece of paper each time a behavior occurs. At the end
of the observation period, the frequency of the behavior is equal to the
number of tears in the paper (Epstein, 1996).

â–  Use Ranger beads. Ranger beads (brought to my attention by Jason Hicks, a
student in my behavior modification class, who first used them when he
was an Army Ranger) consist of a strip of leather or nylon threaded through
beads. They have two sections, each with nine beads. With the beads in
one section, the person can record 1 through 9; with the beads in the
other section, the person can count by 10s, for a maximum frequency
count of 99. Whenever a target behavior occurs, the person moves a bead
from one side of the strip to the other. At the end of the day or observation
period, the number of beads moved to one side indicates the frequency of
the target behavior. A similar recording system involves beads on a piece of
leather or string worn around the wrist.

â–  Use a pedometer. The pedometer is an automatic device, worn on the belt,
that records each step a person takes while walking or running.

Regardless of the instrument used, the one characteristic of all behavior recording
procedures is that the person observes the behavior and records it immediately
(the exception is when a device such as a pedometer automatically records the
behavior). The sooner the observer records the behavior after it occurs, the less
likely the observer is to record incorrectly. A person who waits some time to
record an observation may forget to record it at all.

One other aspect of a behavior recording procedure is that it must be practi-
cal. The person responsible for recording the target behavior must be able to use
the recording procedure without much difficulty or disruption of ongoing activities.
If a recording procedure is practical, the person is more likely to carry out the
recording (or self-monitoring) successfully. A recording procedure that takes sub-
stantial time or effort is not practical. In addition, the recording procedure should
not draw attention to the person who is doing the observation and recording. If
this happens, the person may be less likely to carry out the recording procedure.

Reactivity

Sometimes the process of recording a behavior causes the behavior to change,
even before any treatment is implemented. This is called reactivity (Foster, Bell-
Dolan, & Burge, 1988; Hartmann & Wood, 1990; Tryon, 1998). Reactivity may
occur when an observer is recording the behavior of another person or when a
person engages in self-monitoring. Reactivity may be undesirable, especially for
research purposes, because the behavior recorded during the observation period is
not representative of the level of the behavior occurring in the absence of the

32 Chapter 2

observer or in the absence of self-monitoring. For example, when a disruptive
child sees that someone is recording his or her behavior in the classroom, the
child may decrease his or her disruptive behavior while the observer is present.
Usually this change in behavior is only temporary, and the behavior returns to its
original level once the child becomes accustomed to the observer’s presence.

One way to reduce reactivity is to wait until the people who are being
observed become accustomed to the observer. Another is to have the observer
record the behavior without the people knowing that they are being observed.
This may be accomplished with the use of one-way observation windows or with
participant observers. A participant observer is a person who is normally in the set-
ting where the target behavior occurs, such as a teacher’s aide in a classroom.

Likewise, when a person starts to record his or her own behavior as part of a
self-management project, the behavior often changes in the desired direction as a
result of the self-monitoring (Epstein, 1996). For this reason, self-monitoring
sometimes is used as a treatment to change a target behavior. For example, Ollen-
dick (1981) and Wright and Miltenberger (1987) found that self-monitoring of
motor tics led to reductions in their frequency. Ackerman and Shapiro (1984)
reported that when adults with intellectual disabilities self-monitored their work
productivity, their productivity increased. Winett, Neale, and Grier (1979) showed
that self-monitoring of electricity use by people in their homes resulted in
decreases in electricity use. Self-monitoring and other self-control strategies are
discussed in more detail in Chapter 20.

Interobserver Agreement

You assess IOA to determine whether the target behavior is being recorded consis-
tently. To evaluate IOA, two people independently observe and record the same
target behavior of the same subject during the same observation period. The
recordings of the two observers are then compared, and a percentage of agreement
between observers is calculated. When the percentage of agreement is high, it
indicates that there is consistency in the scoring by the two observers. This sug-
gests that the definition of the target behavior is clear and objective, and that the
observers are using the recording system correctly. When high IOA is reported in
a research study, it suggests that the observers in the study recorded the target
behavior consistently. IOA should be checked at least occasionally when direct
observation and recording are also used in nonresearch settings. In research studies,
the minimally acceptable IOA is typically 80%, although 90% or better is preferred.

IOA is calculated differently depending on the recording method used. For
frequency recording, IOA (expressed as a percentage) is calculated by dividing
the smaller frequency by the larger frequency. For example, if observer A records
10 occurrences of aggressive behavior in an observation period and observer B
records 9, IOA equals 90%. For duration recording, IOA is calculated by dividing
the smaller duration by the larger duration. For example, if observer A records 48
minutes of exercise and observer B records 50 minutes, IOA equals 48/50, or 96%.
For interval recording, you check the agreement between the two observers in
each interval. You then divide the number of intervals with agreement by the

Observing and Recording Behavior 33

total number of intervals. Agreement is defined as the case in which the two
observers both recorded the target behavior as occurring or as not occurring in a
particular interval. Figure 2-7 shows the interval recording data from two indepen-
dent observers recording the behavior of the same client at the same time. There
were 20 intervals of observation and the 2 observers agreed on the occurrence or
nonoccurrence of the behavior 17 times. Therefore, you divide 17 by 20, which
equals 0.85, or 85%. IOA for time sample recording is calculated in the same
manner as for interval recording.

There are two variations of IOA calculations for interval recording; occur-
rence only IOA and nonoccurrence only IOA. In occurrence only IOA, only
those intervals where both observers scored an occurrence of the behavior are
counted as agreements. Intervals where both observers did not score an occur-
rence of the behavior are not used in the calculation. In nonoccurrence only
IOA, only those intervals where both observers agreed the behavior did not occur
are counted as agreements. Intervals where both observers scored an occurrence of
the behavior are not used in the calculation. Occurrence only IOA calculations
provide a more conservative measure of IOA for low rate behaviors because it is
easy to agree on the nonoccurrence of the behavior by chance. Nonoccurrence
only IOA calculations provide a more conservative measure of IOA for high rate
behaviors because it is easier to agree on the occurrence of the behavior by
chance. Figure 2-8 shows occurrence only IOA calculation and Figure 2-9 shows
nonoccurrence only IOA calculation.

Observer A

A A A A A A DA A A A A AAD

X X X X X X X XX

A D A AA

XX

Observer B X X X X X X XX X XX X

A /( A + D) = 17/20 = 0.85 = 85%

FIGURE 2-7 A comparison of interval recording by two observers. An A indicates that the observers agreed that
the behavior did or did not occur in an interval. D indicates that the observers disagreed: One
observer recorded the occurrence of the behavior in an interval, and the other did not.

Observer A

A AD A

X X X X X

Observer B X X X X

A

A /( A + D) = 4/5 = 80%

FIGURE 2-8 Calculation of IOA using agreement on occurrence only. The number of intervals with agreements
on the occurrence of the behavior is divided by intervals with agreements + disagreements. Inter-
vals where both observers did not score an occurrence are not included in the calculation.

34 Chapter 2

To calculate IOA for frequency-within-interval recording, you calculate a percentage
of agreement between observers for each interval (smaller frequency divided by the
larger frequency), sum the percentages for all the intervals, and divide by the number
of intervals in the observation period. Figure 2-10 shows the calculation of IOA for two
independent observers using frequency-within-interval recording.

CHAPTER SUMMARY

1. A target behavior is defined by identifying exactly
what the person says or does that constitutes the
behavioral excess or behavioral deficit targeted for
change. The behavioral definition should include
active verbs describing the behavior the person
exhibits.

2. The different methods you can use to record the
target behavior include continuous recording of
the frequency, duration, latency, or magnitude
of the behavior; percentage of opportunity record-
ing; product recording; interval recording; or time
sample recording.

3. With continuous recording, the observer observes
the client continuously throughout the observa-
tion period and records each occurrence of the
behavior. With interval and time sample record-
ing, the observation period is divided into a num-
ber of smaller time periods or intervals and the
behavior is recorded as occurring or not occurring
within each interval. With interval recording, the
intervals are consecutive time periods, and with
time sample recording, the intervals are separated
by periods without observation.

Observer A

D A A A A

X X X X X X X X XXXX XX

Observer B X X X X X X X X X XXXX XX

A

A /( A + D) = 5/6 = 83.3%

FIGURE 2-9 Calculation of IOA using agreement on nonoccurrence only. The number of intervals with agree-
ments on the nonoccurrence of the behavior is divided by intervals with agreements + disagree-
ments. Intervals where both observers scored an occurrence are not included in the calculation.

Observer A XXX X XX XXXX XXX X XX XXX

Observer B XXX X XXX XXX X X XXX XXX

3/3 1/1 2/3 0/0 3/4 1/3 0/0 1/1 2/3 3/3

100% + 100% + 67% + 100% + 75% + 33% + 100% + 100% + 67% + 100% = 842%
842% divided by 10 (the number of intervals) = 84.2%

FIGURE 2-10 Calculation of interobserver agreement for frequency-within-interval recording. A percentage of
agreement is calculated for each interval, the percentages are summed, and the sum is divided by
the number of intervals.

Observing and Recording Behavior 35

4. Reactivity occurs when the process of behavior
recording causes the behavior to change, even
before any treatment is implemented. Reactivity
can be minimized by waiting until the person
being observed becomes accustomed to the obser-
ver’s presence. Another way to reduce reactivity is
to observe people without letting them know they
are being observed.

5. Interobserver agreement (IOA) is determined
by having two observers independently record a
person’s behavior during the same observation
period and then comparing the recordings of
the two observers. You assess IOA to determine
whether the target behavior is being recorded
consistently.

KEY TERMS

analogue setting, 22
baseline, 25
behavioral assessment, 17
continuous recording, 23
direct assessment, 18
duration, 23
frequency, 23
frequency-within-interval

recording, 28

indirect assessment, 17
intensity, 24
interobserver agreement, 19
interobserver reliability, 19
interval recording, 27
latency, 24
natural setting, 22
observation period, 21
product recording, 27

rate, 23
reactivity, 32
real-time recording, 24
self-monitoring, 21
structured observation, 22
time sample recording, 28
unstructured observation, 22

PRACTICE TEST

1. Why is it important to record the behavior you
are trying to change when using behavior modi-
fication? (p. 17)

2. Identify the four steps involved in a behavior
recording plan. (p. 19)

3. What is a behavioral definition? How does it
differ from a label for a behavior? (p. 19)

4. Provide a possible behavioral definition of
politeness.

5. Why is it important to identify who will record a
behavior? (p. 21)

6. What is meant by the term observation period?
(p. 21)

7. Identify and define four dimensions of a behav-
ior that may be recorded in a continuous record-
ing method. (pp. 23–24)

8. Provide an example of frequency recording,
duration recording, intensity recording, and
latency recording. (pp. 23–24)

9. What is real-time recording? Provide an exam-
ple. (p. 24)

10. What is product recording? Provide an example.
(p. 27)

11. What is interval recording? Provide an example.
(p. 27)

12. What is frequency-within-interval recording?
Provide an example. (p. 28)

13. What is time sample recording? Provide an
example. (p. 28)

14. Provide examples of three different recording
instruments. (pp. 29–32)

15. Why is it important to record a behavior imme-
diately after it occurs? (p. 32)

16. What is reactivity? Describe two ways to
reduce reactivity during direct observation.
(pp. 32–33)

17. What is interobserver agreement, and why is it
assessed? (p. 33)

18. Describe how you calculate interobserver agree-
ment for frequency recording, duration record-
ing, and interval recording. (pp. 33–34)

19. Describe how you would calculate interobserver
agreement for frequency-within-interval record-
ing. (p. 35)

36 Chapter 2

APPLICATIONS

1. When people want to change their own behav-
ior, they can design and implement a self-
management program. A self-management pro-
gram involves applying behavior modification to
one’s own behavior. There are five steps in a self-
management program:

i. Self-monitoring. Define and record the tar-
get behavior you want to change.

ii. Graphing. Develop a graph and plot the
daily level of the target behavior on the
graph.

iii. Goal setting. Establish a goal for the desired
change in the target behavior.

iv. Intervention. Develop and implement spe-
cific behavior modification strategies to
change the behavior.

v. Evaluation. Continue to record the behav-
ior and plot it on the graph to determine
whether you changed your target behavior
and achieved your goal.

In this exercise, take the first step to start
your own self-management program. Define a
target behavior you want to change, and develop
a behavior recording plan to measure the target
behavior. As you complete this first step, con-
sider the following questions:

a. Did you define your target behavior in clear,
objective terms?

b. Did you determine the appropriate dimen-
sion of your target behavior to record (e.g.,
frequency or duration)?

c. Did you choose a practical recording
method?

d. Will you be able to record your target behavior
immediately each time that it occurs?

e. What problems might you encounter as you
record your target behavior, and how will you
deal with these problems?

Good luck in starting the self-monitoring
component of your self-management program.
You will learn the information you need to

conduct the remaining steps of your self-
management program in subsequent chapters.

2. Imagine you have a friend, James, who is study-
ing to be an elementary school teacher. James is
doing his student teaching this semester in a
second grade classroom in a public school.
James mentioned to you that one of his students
has trouble staying in her seat, paying attention
during class, and participating in activities. This
student, Sara, gets out of her seat and talks to or
teases other children. When she is out of her
seat, she does not pay attention to James, does
not participate in activities, and disrupts the
other children.

James believes that if he could just get Sara
to stay in her seat, he could get her to pay atten-
tion and participate. As a result, she would per-
form better in class, and the other students
would also do better. James knows you are tak-
ing a behavior modification class, so he has
come to you for help.

You inform James that the first step he must
take, if he is going to use behavior modification
with Sara, is to develop a recording plan to mea-
sure her behavior. In this exercise, develop
a plan that James could use to record Sara’s
out-of-seat behavior. Consider the following
questions:

a. What is the behavioral definition of out-
of-seat behavior?

b. What recording method will you have James
use to record Sara’s out-of-seat behavior?

c. What instrument will you have James use to
record the behavior? Will this instrument be
practical for James to use as a teacher?

3. Eve plans to start a weight-lifting program. She
wants to record her behavior once she starts the
program so that she can measure the changes in
her behavior as the program progresses.
Describe how Eve could use frequency record-
ing, duration recording, and intensity recording
to measure her weight-lifting behavior.

Observing and Recording Behavior 37

MISAPPLICATIONS

1. Gloria is taking a behavior modification class
and has to do a self-management project. The
behavior she has chosen to modify is her hair-
twirling. She has defined this behavior as any
instance in which she reaches up to the back
of her head and wraps hair around her finger.
The first step in her self-management project is
to develop a behavior recording plan. Because
she usually does the hair-twirling in her classes,
she decides to record the behavior immediately
after each class period. She will keep a 3�5-inch
note card in her purse, and as soon as she leaves
the classroom, she will get the note card out of
her purse and write down the number of times
that she twirled her hair in the class.
a. What is wrong with this behavior recording

plan?
b. What changes would you make to improve

it?
2. Ralph is going to implement a self-management

project to help him decrease the number of
cigarettes he smokes per day. He will define
the behavior of smoking a cigarette as any
instance in which he takes a cigarette out of
the pack in his pocket, lights it, and smokes
any part of it. He will record the number of

cigarettes he smokes each day by counting the
cigarettes left in his pack at the end of the day
and subtracting this number from the number of
cigarettes that were in the pack at the start of the
day.
a. What is wrong with this behavior recording

plan?
b. What would you do to improve it?

3. Below are examples of behavioral definitions of
target behaviors in students’ self-management
programs. What is wrong with each of these
behavioral definitions?
a. Losing my temper will be defined as getting

mad at my husband and yelling at him,
walking into the bedroom and slamming the
door, or telling him to “shut up” when he
says something that frustrates me.

b. Overeating will be defined as any time I eat
more than I wanted to eat at a meal, or any
time I eat so much that I feel bloated or my
belt is too tight.

c. Studying will be defined as any time I have
my books open in front of me in the library or
at my desk, the TV is off, and there are no
other distractions.

38 Chapter 2

Chapter Three

Graphing Behavior and Measuring
Change

As we saw in Chapter 2, people who use behavior modification define theirtarget behavior carefully, and directly observe and record the behavior. In
this way, they can document whether the behavior has indeed changed when a
behavior modification procedure is implemented. The primary tool used to docu-
ment behavior change is the graph.

A graph is a visual representation of the occurrence of a behavior over time.
After instances of the target behavior are recorded (on a data sheet or otherwise),
the information is transferred to a graph. A graph is an efficient way to view the
occurrence of the behavior because it shows the results of recording during many
observation periods.

Behavior analysts use graphs to identify the level of behavior
before treatment and after treatment begins. In this way, they can
document changes in the behavior during treatment and make deci-
sions about the continued use of the treatment. The graph makes it
easier to compare the levels of the behavior before, during, and after
treatment because the levels are presented visually for comparison.
In Figure 3-1, for example, it is easy to see that the frequency of
the behavior is much lower during treatment (competing response)
than before treatment (baseline). This particular graph is from a stu-
dent’s self-management project. The student’s target behavior
involved biting the insides of her mouth when she studied. She
recorded the behavior on a data sheet each time it occurred. After
10 days of recording the behavior without any treatment (baseline),
she implemented a behavior modification plan in which she used a
competing response (a behavior that is incompatible with mouth-
biting and interrupts each occurrence of mouth-biting) to help her

control the mouth-biting behavior. After implementing this competing response
procedure, she continued to record the behavior for 20 more days. She then
recorded the behavior four more times, after 1, 5, 10, and 20 weeks. The long
period after treatment has been implemented is called the follow-up period.
From this graph, we can conclude that the mouth-biting behavior (as recorded
by the student) decreased substantially while the student implemented the

â–  What are the six essential components
of a behavior modification graph?

â–  How do you graph behavioral data?

â–  What different dimensions of behavior
can be shown on a graph?

â–  What is a functional relationship, and
how do you demonstrate a functional
relationship in behavior modification?

â–  What different research designs can
be used in behavior modification
research?

39

treatment. We can also see that the behavior continued to occur at a low level up
to 20 weeks after treatment was implemented.

Components of a Graph

In the typical behavior modification graph, time and behavior are the two vari-
ables illustrated. Each data point on a graph gives you two pieces of information:
It tells you when the behavior was recorded (time) and the level of the behavior at
that time. Time is indicated on the horizontal axis (also called the x-axis, or the
abscissa), and the level of the behavior is indicated on the vertical axis (also called
the y-axis, or the ordinate). In Figure 3-1, the frequency of mouth-biting is indi-
cated on the vertical axis, and days and weeks are indicated on the horizontal
axis. By looking at this graph, you can determine the frequency of mouth-biting
on any particular day, before or after treatment was implemented. Because
follow-up is reported, you can also see the frequency of the behavior at intervals
of up to 20 weeks.

Six components are necessary for a graph to be complete.

â–  The y-axis and the x-axis. The vertical axis (y-axis) and the horizontal axis
(x-axis) meet at the bottom left of the page. On most graphs, the x-axis is longer
than the y-axis; it is usually one to two times as long (Figure 3-2).

â–  The labels for the y-axis and the x-axis. The y-axis label usually tells you
the behavior and the dimension of the behavior that is recorded. The x-axis
label usually tells you the unit of time during which the behavior is recorded.
In Figure 3-3, the y-axis label is “Hours of Studying” and the x-axis label is
“Days.” Thus, you know that the hours of studying will be recorded each day
for this particular person.

Fr
eq

ue
nc

y
of

M
ou

th
B

iti
ng

0
5

10

20
25

15

30
35

45
40

50
55
60
65

75
70

80

50 10 15 20 25 201051
Days Weeks

Baseline Competing Response Follow-Up

••

•

•

•

•

•

••

•

•

•
•

••••••
•

••••••••• • • • •

FIGURE 3-1 This graph shows the frequency of mouth-biting during baseline and treatment (competing
response) phases and follow-up.

40 Chapter 3

â–  The numbers on the y-axis and the x-axis. On the y-axis, the numbers indi-
cate the units of measurement of the behavior; on the x-axis, the numbers indicate
the units of measurement of time. There should be a hash mark on the y-axis
and the x-axis to correspond to each of the numbers. In Figure 3-4, the numbers
on the y-axis indicate the number of hours the studying behavior occurred, and
the numbers on the x-axis indicate the days on which studying was measured.

y
-a

xi
s

x-axis

FIGURE 3-2 The y-axis and the x-axis.

Days

H
ou

rs
o

f S
tu

dy
in

g

FIGURE 3-3 Labels for the y-axis and the x-axis.

H
ou

rs
o

f S
tu

dy
in

g

0

5

1

2

4

3

20 4 6 8 10 12 14 16 18 20
Days

FIGURE 3-4 Numbers on the y-axis and the x-axis.

Graphing Behavior and Measuring Change 41

â–  Data points. The data points must be plotted correctly to indicate the level
of the behavior that occurred at each particular time period. The information on
the level of the behavior and the time periods is taken from the data sheet or other
behavior-recording instrument. Each data point is connected to the adjacent data
points by a line (Figure 3-5).

â–  Phase lines. A phase line is a vertical line on a graph that indicates a change in
treatment. The change can be from a no-treatment phase to a treatment phase, from
a treatment phase to a no-treatment phase, or from one treatment phase to another
treatment phase. A phase is a period in which the same treatment (or no treatment)
is in effect. In Figure 3-6, the phase line separates baseline (no treatment) and treat-
ment phases. Data points are not connected across phase lines. This allows you to see
differences in the level of the behavior in different phases more easily.

â–  Phase labels. Each phase in a graph must be labeled. The phase label
appears at the top of the graph above the particular phase (Figure 3-7). Most
behavior modification graphs have at least two phases that are labeled: the no-
treatment phase and the treatment phase. “Baseline” is the label most often
given to the no-treatment phase. The label for the treatment phase should identify
the particular treatment being used. In Figure 3-7, the two phase labels are “Base-
line” and “Behavioral Contract.” The behavioral contract is the particular treat-
ment the student is using to increase studying. Some graphs have more than one
treatment phase or more than one baseline phase.

H
ou

rs
o

f S
tu

dy
in

g
0

5

1

2

4

3

20 4 6 8 10 12 14 16 18 20
Days

• •
• •

• • •

•
• • • • • • • • •

• • •

FIGURE 3-5 Data points plotted on a graph.

• •
• •

• • •

•
• • • • • • • • •

• • •

H
ou

rs
o

f S
tu

dy
in

g

0

5

1

2

4

3

20 4 6 8 10 12 14 16 18 20
Days

FIGURE 3-6 Phase line on a graph.

42 Chapter 3

Graphing Behavioral Data

As discussed in Chapter 2, behavioral data are collected through direct observation
and recording of the behavior on a data sheet or other instrument. Once the
behavior has been recorded on the data sheet, it can be transferred to a graph.
For example, Figure 3-8a is a frequency data sheet that shows 2 weeks of behavior
recording, and Figure 3-8b is a graph of the behavioral data from the data sheet.
Notice that days 1-14 on the data sheet correspond to the 14 days on the graph.
Also notice that the frequency of the behavior listed on the data sheet for each
day corresponds to the frequency recorded on the graph for that day. As you look
at the graph, you can immediately determine that the frequency of the behavior is
much lower during treatment than during baseline. You have to look more closely
at the data sheet to be able to detect the difference between baseline and treat-
ment. Finally, notice that all six essential components of a graph are included in
this graph.

Consider a second example. A completed duration data sheet is shown in Fig-
ure 3-9a, and Figure 3-9b is a table that summarizes the daily duration of the
behavior recorded on the data sheet. Notice that the duration of the behavior
listed in the summary table for each of the 20 days corresponds to the duration
that was recorded each day on the data sheet.

Below the data summary table (see Figure 3-9b) is a graph that is only partially
completed (see Figure 3-9c). Using the information provided in the data sum-
mary table, complete this graph. Be sure that the completed graph includes all
six components discussed earlier.

To complete Figure 3-9c, you must add four components. First, you should add
the data points for days 8-20 and connect them. Second, include the phase line
between days 7 and 8. Data points on days 7 and 8 should not be connected across
the phase line. Third, add the phase label “Behavioral Contract,” to the right of the
phase line. Fourth, add the label “Days” to the x-axis. When these four components
are added, the graph includes all six essential components (Figure 3-10).

• •
• •

• • •

•
• • • • • • • • •

• • •

Baseline Behavioral Contract

H
ou

rs
o

f S
tu

dy
in

g

0

5

1

2

4

3

20 4 6 8 10 12 14 16 18 20
Days

FIGURE 3-7 Phase labels on a graph.

Graphing Behavior and Measuring Change 43

FOR FURTHER READING
Graphing in Excel

Although it is easy to construct a graph with a piece of graph paper, a ruler, and a pencil, there are
graphing programs that allow you to construct a graph on your computer. Graphs can be constructed
in two different Microsoft Office programs; PowerPoint and Excel. Carr and Burkholder (1998) and
Dixon et al. (2007) published articles in the Journal of Applied Behavior Analysis providing step-by-step
instructions for how to use Microsoft Excel for constructing the types of graphs used in applied behavior
analysis or behavior modification. Students interested in learning how to construct graphs in Excel are
encouraged to read these articles.

Days

1

Ci
ga

re
tt

es
S

m
ok

ed

0

2

4

6

10

8

20 4 8 106 1412

• •
• •

•
•

• •
• •

• •
• •

Baseline Behavioral Contract

*Day 6 was the last day of baseline and day 7 was the first day of treatment.

Days

Frequency(a)

(b)

1
2
3
4
5
6*
7
8
9
10
11
12
13
14

8
8
7
7
9
8
5
5
4
4
3
3
2
2

Daily Total2 3 4 5 6 9 10 11 127 8

X
X
X
X
X
X
X
X

X
X
X
X
X
X

X
X
X
X
X
X
X
X

X
X
X
X
X
X

X
X
X
X
X
X
X
X

X
X

X
X
X
X
X
X
X
X

X
X
X
X
X
X

X
X
X
X
X
X

X
X
X
X
X
X

X
X

X X
X

X
X

X
X

X
X

FIGURE 3-8 A completed frequency data sheet is shown in (a); the number of cigarettes smoked each day is
recorded on the sheet. The graph of the behavioral data from the data sheet (b) is also shown.
The treatment involved a behavioral contract in which the client agreed to smoke one fewer ciga-
rette per day every second day. Behavioral contracts are described in Chapter 23.

44 Chapter 3

Onset

M
in

ut
es

o
f E

xe
rc

is
e

0

15

30

45

60

20 4 8 106 1412 16 2018
•• • •

• •

Baseline

*Baseline ended on day 7. On day 8, the subject implemented treatment involving a behavioral contract.

Days

(a)

(b)

1
2
3
4
5
6
7*
8
9
10
11
12
13
14

0
15
0
0

15
0
0

15
30
30
15
30
0

30
15
16
17
18
19
20

Days 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20

0 15 0 0 15 0 0 15 30 30 15 30 0 30 45 30 45 45 0 60Duration
(minutes)

45
30
45
45
0

60

Daily DurationOffset Onset Onset OffsetOffset

7:00 7:15

7:10 7:25

7:30 8:00
7:30 8:00

7:00 7:15

6:30 6:45
6:45 7:15

7:00 7:30

6:45 7:15
6:30 6:45

7:00 7:30

6:45 7:15

7:00 7:30

7:45 8:00

7:30 8:00

6:30 7:15

(c)

• •

FIGURE 3-9 A completed duration data sheet is shown (a); the sheet is used to record the duration of exercise each
day. The completed data summary table (b) is also shown. The incomplete graph (c) is for the student to
complete using the behavioral data in (b).

Graphing Behavior and Measuring Change 45

Graphing Data from Different Recording
Procedures

Figures 3-8 and 3-10 illustrate graphs of frequency data and duration data, respec-
tively. Because other types of data can be recorded, other types of graphs are possi-
ble. Regardless of the dimension of behavior or type of data that is being graphed,
however, the same six components of a graph must be present. What will change
with different recording procedures are the y-axis label and the numbering on the
y-axis. For example, if you are recording the percentage of math problems a stu-
dent completes correctly during each math class, you would label the y-axis “Per-
centage of Correct Math Problems” and number the y-axis from 0% to 100%. As
you can see, the y-axis label identifies the behavior (correct math problems) and
the type of data (percentage) that is recorded.

Consider another example. A researcher is studying Tourette’s syndrome, a neu-
rological disorder in which certain muscles in the body twitch or jerk involuntarily
(these are called motor tics). The researcher uses an interval recording system and
records whether a motor tic occurs during each consecutive 10-second interval in
30-minute observation periods. At the end of each observation period, the researcher
calculates the percentage of intervals in which a tic occurred. The researcher labels
the y-axis of the graph “Percentage of Intervals of Tics” and numbers the y-axis from
0% to 100%. Whenever an interval recording system is used, the y-axis is labeled
“Percentage of Intervals of (Behavior).” The x-axis label indicates the time periods in
which the behavior was recorded (e.g., “Sessions” or “Days”). The x-axis is then num-
bered accordingly. A session is a period in which a target behavior is observed and
recorded. Once treatment is started, it is also implemented during the session.

Other aspects of a behavior may be recorded and graphed, such as intensity or
product data. In each case, the y-axis label should clearly reflect the behavior and
the dimension or aspect of the behavior that is recorded. For example, as a mea-
sure of how intense or serious a child’s tantrums are, you might use the label
“Tantrum Intensity Rating” and put the numbers of the rating scale on the y-axis.
For a measure of loudness of speech, the y-axis label might be “Decibels of
Speech,” with decibel levels numbered on the y-axis. To graph product recording

Days

M
in

ut
es

o
f E

xe
rc

is
e

0

15

30

45

60

20 4 8 106 1412 16 2018
•• • •

•

• •

•

•

•

•

•

•

• •

•

•

• •

• •

Baseline Behavioral Contract

FIGURE 3-10 Completed graph using data from the data summary table in Figure 3-9b.

46 Chapter 3

data, you would label the y-axis to indicate the unit of measurement and the
behavior. For example, “Number of Brakes Assembled” is a y-axis label that indi-
cates the work output of a person who puts together bicycle brakes.

Research Designs

When people conduct research in behavior modification, they use research designs
that include more complex types of graphs. The purpose of a research design is to
determine whether the treatment (independent variable) was responsible for the
observed change in the target behavior (dependent variable) and to rule out the
possibility that extraneous variables caused the behavior to change. In research, an
independent variable is what the researcher manipulates to produce a change in
the target behavior. The target behavior is called the dependent variable. An extra-
neous variable, also called a confounding variable, is any event that the researcher
did not plan that may have affected the behavior. For a person with a problem, it
may be enough to know that the behavior changed for the better after using behav-
ior modification procedures. However, a researcher also wants to demonstrate that
the behavior modification procedure is what caused the behavior to change.

When a researcher shows that a behavior modification procedure causes a tar-
get behavior to change, the researcher is demonstrating a functional relationship
between the procedure and the target behavior. That is, the researcher demon-
strates that the behavior changes as a function of the procedure. A functional rela-
tionship is established if (a) a target behavior changes when an independent
variable is manipulated (a procedure is implemented), while all other variables
are held constant, and (b) the process is replicated or repeated one or more times
and the behavior changes each time. A behavior modification researcher uses a
research design to demonstrate a functional relationship. A research design
involves both treatment implementation and replication. If the behavior changes
each time the procedure is implemented and only when the procedure is imple-
mented, a functional relationship is demonstrated.

In this case, we would say that the researcher has demonstrated experimental
control over the target behavior. It is unlikely that an extraneous variable caused
the behavior change if it changed only when the treatment was implemented.
This section reviews research designs used in behavior modification (for further
information on behavior modification research designs, see Bailey, 1977; Barlow &
Hersen, 1984; Gast, 2009; Hayes, Barlow, & Nelson-Gray, 1999; Kazdin, 2010;
Poling & Grossett, 1986).

A-B Design
The simplest type of design used in behavior modification has just two phases: base-
line and treatment. This is called an A-B design, where A ¼ baseline and B ¼
treatment. A-B designs are illustrated in Figures 3-1, 3-7, 3-8b, and 3-10. By means
of an A-B design, we can compare baseline and treatment to determine whether
the behavior changed in the expected way after treatment. However, the A-B design
does not demonstrate a functional relationship because treatment is not replicated
(implemented a second time). Therefore, the A-B design is not a true research

Graphing Behavior and Measuring Change 47

design; it does not rule out the possibility that an extraneous variable was responsi-
ble for the behavior change. For example, although mouth-biting decreased when
the competing response treatment was implemented in Figure 3-1, it is possible
that some other event (extraneous variable) occurred at the same time as treatment
was implemented. In that case, the decrease in mouth-biting may have resulted
from the other event or a combination of treatment and the other event. For exam-
ple, the person may have seen a TV show about controlling nervous habits and
learned from that how to control her mouth-biting.

Because the A-B design does not rule out other causes, it is rarely used by
behavior modification researchers. It is most often used in applied, nonresearch
situations, in which people are more interested in demonstrating that behavior
change has occurred than in proving that the behavior modification procedure
caused the behavior change. You probably would use an A-B graph in a self-
management project to show whether your behavior changed after you implemen-
ted a behavior modification procedure.

A-B-A-B Reversal Design
The A-B-A-B reversal design is an extension of the simple A-B design (where A ¼
baseline and B ¼ treatment). In the A-B-A-B design, baseline and treatment
phases are implemented twice. It is called a reversal design because after the first
treatment phase, the researcher removes the treatment and reverses back to base-
line. This second baseline is followed by replication of the treatment. Figure 3-11
illustrates an A-B-A-B design.

Sessions

Bob

N
um

be
r o

f A
gg

re
ss

iv
e

Re
sp

on
se

s

0

20

40

60

80

100

120

140

160

180

200

0 5 10 15 20 25

Demands
No

Demands Demands No Demands

••
•

••••

•
•

•
•

•
•

•
••

•

••••••••

FIGURE 3-11 This A-B-A-B graph shows the frequency of aggressive behaviors by an adolescent with intellectual
disability during baseline phases involving demands and treatment phases involving no demands.
(From Carr, E. G., Newsom, C. D., & Binkoff, J. A. [1980]. Escape as a factor in the aggressive
behavior of two retarded children. Journal of Applied Behavior Analysis, 13, 101–117. Copyright
© 1980 University of Kansas Press. Reprinted by permission of the author and the publisher.)

48 Chapter 3

The A-B-A-B graph in Figure 3-11 shows the effect of a teacher’s demands on
the aggressive behavior of an adolescent with intellectual disability named Bob. Carr
and his colleagues (Carr, Newsom, & Binkoff, 1980) studied the influence of
demands on Bob’s aggressive behavior by alternating phases in which teachers made
frequent demands with phases in which teachers made no demands. In Figure 3-11,
you can see that the behavior changed three times. In the baseline phase
(“Demands”), the aggressive behavior occurred frequently. When the treatment
phase (“No Demands”) was first implemented, the behavior decreased. When the
second “Demands” phase occurred, the behavior returned to its level during the first
“Demands” phase. Finally, when the “No Demands” phase was implemented a sec-
ond time, the behavior decreased again. The fact that the behavior changed three
times, and only when the phase changed, is evidence that the change in demands
(rather than some extraneous variable) caused the behavior change. When the
demands were turned on and off each time, the behavior changed accordingly. It is
highly unlikely that an extraneous variable was turned on and off at exactly the same
time as the demands, so it is highly unlikely that any other variable except the inde-
pendent variable (change in demands) caused the behavior change.

Variations of the A-B-A-B reversal design may be used in which more than one
treatment is evaluated. Suppose for example, you implemented one treatment (B)
and it did not work, so you implemented a second treatment (C) and it did work.
To replicate this treatment and show experimental control, you might use an
A-B-C-A-C design. If the second treatment (C) resulted in a change in the target
behavior each time it was implemented, you are demonstrating a functional rela-
tionship between this treatment and the behavior.

A number of considerations must be taken into account in deciding whether
to use the A-B-A-B research design. First, it may not be ethical to remove the treat-
ment in the second baseline if the behavior is dangerous (e.g., self-injurious
behavior). Second, you must be fairly certain that the level of the behavior will
reverse when treatment is withdrawn. If the behavior fails to change when the
treatment is withdrawn, a functional relationship is not demonstrated. Another
consideration is whether you can actually remove the treatment after it is imple-
mented. For example, if the treatment is a teaching procedure and the subject
learns a new behavior, you cannot take away the learning that took place. (For a
more detailed discussion of considerations in the use of the A-B-A-B design, see
Bailey [1977], Bailey and Burch [2002], Barlow and Hersen [1984], Gast [2009],
and Kazdin [2010].)

Multiple-Baseline Design
There are three types of multiple-baseline designs.

â–  In a multiple-baseline-across-subjects design, there is a baseline and a
treatment phase for the same target behavior of two or more different subjects.

â–  In a multiple-baseline-across-behaviors design, there is a baseline and
treatment phase for two or more different behaviors of the same subject.

â–  In a multiple-baseline-across-settings design, there is a baseline and treat-
ment phase for two or more settings in which the same behavior of the same sub-
ject is measured.

Graphing Behavior and Measuring Change 49

Remember that the A-B-A-B design can also have two baseline phases and two
treatment phases, but both baseline and treatment phases occur for the same
behavior of the same subject in the same setting. With the multiple-baseline
design, the different baseline and treatment phases occur for different subjects, or
for different behaviors, or in different settings.

A multiple-baseline design may be used (a) when you are interested in the
same target behavior exhibited by multiple subjects, (b) when you have targeted
more than one behavior of the same subject, or (c) when you are measuring a
subject’s behavior across two or more settings. A multiple-baseline design is useful
when you cannot use an A-B-A-B design for the reasons listed earlier. The
multiple-baseline design and the appropriate time to use it are described in more
detail by Bailey (1977), Bailey and Burch (2002), Barlow and Hersen (1984), Gast
(2009), and Kazdin (2010).

Figure 3-12 illustrates the multiple-baseline-across-subjects design. This
graph, from a study by DeVries, Burnette, and Redmon (1991), shows the effect
of an intervention involving feedback on the percentage of time that emergency
department nurses wore rubber gloves when they had contact with patients.
Notice that there is a baseline and treatment phase for four different subjects
(nurses). Figure 3-12 also illustrates a critical feature of the multiple-baseline
design: The baselines for each subject are of different lengths. Treatment is
implemented for subject 1 while subjects 2, 3, and 4 are still in baseline. Then,
treatment is implemented for subject 2 while subjects 3 and 4 are still in base-
line. Next, treatment is implemented for subject 3 and, finally, for subject 4.
When treatment is implemented at different times, we say that treatment is stag-
gered over time. Notice that the behavior increased for each subject only after
the treatment phase was started for that subject. When treatment was implemen-
ted for subject 1, the behavior increased, but the behavior did not increase at
that time for subjects 2, 3, and 4, who were still in baseline and had not yet
received treatment. The fact that the behavior changed for each subject only
after treatment started is evidence that the treatment, rather than an extraneous
variable, caused the behavior change. It is highly unlikely that an extraneous var-
iable would happen to occur at exactly the same time that treatment started for
each of the four subjects.

A multiple-baseline-across-behaviors design is illustrated in Figure 3-13. This
graph, from a study by Franco, Christoff, Crimmins, and Kelly (1983), shows the
effect of treatment (social skills training) on four different social behaviors of a shy
adolescent: asking questions, acknowledging other people’s comments, making eye
contact, and showing affect (e.g., smiling). Notice in this graph that treatment is
staggered across the four behaviors, and that each of the behaviors changes only
after treatment is implemented for that particular behavior. Because each of the
four behaviors changed only after treatment was implemented for that behavior,
the researchers demonstrated that treatment, rather than some extraneous variable,
was responsible for the behavior change.

A graph used in a multiple-baseline-across-settings design would look like those
in Figures 3-12 and 3-13. The difference is that in a multiple-baseline-across-settings
graph, the same behavior of the same subject is being recorded in baseline and
treatment phases in two or more different settings, and treatment is staggered across
the settings.

50 Chapter 3

Pe
rc

en
ta

ge
o

f C
on

ta
ct

S
itu

at
io

ns
w

ith
G

lo
ve

s
W

or
n

0

100

20

40

80

60

0 5 10 15
Consecutive 10-Hour Shifts

Nurse 4Nurse 4

Nurse 3Nurse 3

Nurse 2Nurse 2

Nurse 1Nurse 1

• ••
•

•
•

•

•

•
•

• •

•
•

•

0

100

20

40

80

60

•

•

•
•

• •

•
•

• •
•

•

•

•
•

• • •

0

100

20

40

80

60

• •
•

•
•

•
•

••
•

• •
• •

•

0

100

20

40

80

60
• •• •

•

• • •
•

•
•

•

•

• •

Baseline Intervention

Feedback

FIGURE 3-12 This multiple-baseline-across-subjects graph shows the percentage of time that four emergency
department nurses wear rubber gloves when they have contact with patients. The intervention,
which involves feedback from their supervisor, is staggered over time and results in an increase in
the behavior for each of the four nurses. (From DeVries, J. E., Burnette, M. M., & Redmon, W. K.
[1991]. AIDS prevention: Improving nurses’ compliance with glove wearing through performance
feedback. Journal of Applied Behavior Analysis, 24, 705–711. Copyright © 1991 University of Kan-
sas Press. Reprinted by permission of the author and the publisher.)

Graphing Behavior and Measuring Change 51

Ra
tin

g
of

A
ff

ec
t

Se
co

nd
s

of
E

ye
C

on
ta

ct
N

um
be

r o
f S

pe
ec

h
A

ck
no

w
le

dg
er

s/
Re

in
fo

rc
er

s
N

um
be

r o
f C

on
ve

rs
at

io
na

l
Q

ue
st

io
ns

0
0 10 3020

Sessions

• ••• •
•••

• ••• •
••

• • ••• • •
•

•
•

•

•
•

•
•

•
• •

•

•• • •
••

• ••• •

•
•

•

• •
•

•
••

• ••• • ••• • ••• • ••• •
••• • ••• • ••• •

•

•

•

• ••

•

•

•
•

•
•

•

•

•

• •••

•

••• •

•

• ••• • ••• • ••• •

•

•• •

•

•
•

•
•

•
•

•

•

•

•
•

•
•

•

• •

•

•

•

•
•

•

• •

•

•
•

•

•

• • ••
•

•

•

• •••

•

••
• • ••

• •

•

•

•

•
•••

•

•
•

•
•

•

•

•
•

••
•

•

•
•

•

•

•
•

•

•

••

•

•

•
•

• •
•

•
•

•
•

0

0

8

1
2

4
3

5
6
7

8

1
2

4
3

5
6
7

0

40

8

16

32

48

24

250

50

100

200

280

150

5040

Baseline Follow-UpTreatment

FIGURE 3-13 This multiple-baseline-across-behaviors graph shows four social behaviors exhibited by a shy ado-
lescent. A social skills training intervention is applied to each of these four behaviors, and each
behavior increases when the intervention is applied to it. (From Franco, D. P., Christoff, K. A., Crim-
mins, D. B., & Kelly, J. A. [1983]. Social skills training for an extremely shy young adolescent: An
empirical case study. Behavior Therapy, 14, 568–575. Copyright © 1983. Reprinted by permission
of Jeffrey Kelly.)

52 Chapter 3

Draw a graph of a multiple-baseline-across-settings design with hypothetical
data. Be sure to include all six components of a complete graph. Assume that
you have recorded the disruptive behavior of a student in two different class-
rooms using an interval recording system. Include baseline and treatment across
two settings in the graph.

The graph in Figure 3-14, from a study by Dunlap, Kern-Dunlap, Clarke,
and Robbins (1991), shows the percentage of intervals of disruptive behavior by a
student during baseline and treatment (revised curriculum) in two settings, the
morning and afternoon classrooms. It also shows follow-up, in which the research-
ers collected data once a week for 10 weeks. Notice that treatment was implemen-
ted first in one setting and then in the other, and the student’s disruptive behavior
changed only after treatment was implemented in each setting. Your graph of a
multiple-baseline-across-settings design would look like Figure 3-14.

FOR FURTHER READING
Nonconcurrent Multiple-Baseline-Across-Subjects Design

In a multiple-baseline-across-subjects design, data collection starts in each of the baselines (for each of
the subjects) at around the same time and the treatment phase is then staggered across time. However,
in a nonconcurrent multiple baseline (MBL) across subjects design (Carr, 2005; Watson & Workman,
1981) the subjects do not participate in the study concurrently. In a nonconcurrent MBL design, the base-
lines for two or more subjects may begin at different points in time. The nonconcurrent MBL is equivalent
to a number of different A-B designs with each participant having a different baseline length. Treatment is
then staggered across baselines of different lengths rather than across time. As long as each of the sub-
jects has a different number of baseline data points before treatment is implemented, the research design
is considered a nonconcurrent MBL. The advantage of a nonconcurrent MBL is that participants may be
evaluated at different points in time; they may be brought into the study consecutively rather than concur-
rently, which is often more practical for researchers to carry out (Carr, 2005).

Alternating-Treatments Design
The alternating-treatments design (ATD), also called a multielement design, dif-
fers from the research designs just reviewed in that baseline and treatment condi-
tions (or two treatment conditions) are conducted in rapid succession and
compared with each other. For example, treatment is implemented on one day,
baseline the next day, treatment the next day, baseline the next day, and so on. In
the A-B, A-B-A-B, or multiple-baseline designs, a treatment phase occurs after a
baseline phase has been implemented for a period of time; that is, baseline and
treatment occur sequentially. In these designs, a baseline or treatment phase is con-
ducted until a number of data points are collected (usually at least three) and there
is no trend in the data. A trend means the data are increasing or decreasing across a
phase. In the ATD, two conditions (baseline and treatment or two different treat-
ments) occur during alternating days or sessions. Therefore, the two conditions can
be compared within the same time period. This is valuable because any extraneous
variables would have a similar effect on both conditions, and thus an extraneous
variable could not be the cause of any differences between conditions.

Consider the following example of an ATD. A teacher wants to determine
whether violent cartoons lead to aggressive behavior in preschool children.

Graphing Behavior and Measuring Change 53

The teacher uses an ATD to demonstrate a functional relationship between violent
cartoons and aggressive behavior. On one day, the preschoolers do not watch any car-
toons (baseline) and the teacher records the students’ aggressive behavior. The next
day, the students watch a violent cartoon and the teacher again records their aggres-
sive behavior. The teacher continues to alternate a day with no cartoons and a day
with cartoons. After a few weeks, the teacher can determine whether a functional rela-
tionship exists. If there is consistently more aggressive behavior on cartoon days and
less aggressive behavior on no-cartoon days, the teacher has demonstrated a functional
relationship between violent cartoons and aggressive behavior in the preschoolers. An
example of a graph from this hypothetical ATD is shown in Figure 3-15.

In this graph, the number of aggressive behaviors occurring per day is graphed
on days when the children watched violent cartoons (odd-numbered days) and on

••

Pe
rc

en
ta

ge
o

f I
nt

er
va

ls

Days Weekly Probes

•

•

• •

•

• • •• •• ••
•

•

•

• • •

••

• • • •
• •

•

•
•

• • • •••••••••••••••••••••••••••••••••••• • • • • • • • • • •

••••••••••••••••••••••••••• • • • • • • • • • •

•
•

••

• •
•

0

80

90

100

10

20

40

30

50

60

70

Baseline Follow-UpRevised Curriculum

0

80

90

100

10

20

40

30

50

60

70

• •

•

• •
• •

1 5

P.M.

A.M.

FIGURE 3-14 This multiple-baseline-across-settings design shows the effect of a revised curriculum on the dis-
ruptive behavior of an adolescent in a classroom setting in the morning (A.M.) and in another class-
room setting in the afternoon (P.M.). The authors used interval recording and put the percentage
of intervals of disruptive behavior on the graph. (From Dunlap, G., Kern-Dunlap, L, Clarke, S., &
Robbins, F. [1991]. Functional assessment, curricular revision, and severe behavior problems. Jour-
nal of Applied Behavior Analysis, 24, 387–397. Copyright © 1991 Society for the Experimental Anal-
ysis of Behavior. Reprinted with permission of the Society for the Experimental Analysis of Behavior.)

54 Chapter 3

days when they did not (even-numbered days). Notice that the aggressive behavior
occurs more frequently on days when the children watched cartoons. Because the
aggressive behavior is always greater on cartoon days, the researchers conclude
that the aggressive behavior occurred as a function of watching violent cartoons.

Changing-Criterion Design
A changing-criterion design typically includes a baseline and a treatment phase.
What makes a changing-criterion design different from an A-B design is that,
within the treatment phase, sequential performance criteria are specified; that is,
successive goal levels for the target behavior specify how much the target behavior
should change during treatment. The effectiveness of treatment is determined by
whether the subject’s behavior changes to meet the changing performance crite-
ria. That is, does the subject’s behavior change each time the goal level changes?
A graph used in a changing-criterion design indicates each criterion level so that
when the behavior is plotted on the graph, we can determine whether the level
of the behavior matches the criterion level.

Consider the graph in Figure 3-16, from a study by Foxx and Rubinoff
(1979). These researchers helped people reduce their excessive caffeine consump-
tion through a positive reinforcement and response cost procedure. (These proce-
dures are discussed in Chapters 15 and 17.) As you can see in the graph, they set
four different criterion levels for caffeine consumption, each lower than the previ-
ous level. When subjects consumed less caffeine than the criterion level, they
earned money. If they drank more, they lost money. This graph shows that treat-
ment was successful: This subject’s caffeine consumption level was always below
each of the criterion levels. Because the subject’s behavior changed each time
the performance criterion changed, it is unlikely that an extraneous variable was
responsible for the change in behavior. DeLuca and Holborn (1992) used a
changing-criterion design in a study constructed to help obese boys exercise

Fr
eq

ue
nc

y
of

A
gg

re
ss

iv
e

Be
ha

vi
or

0

5

10

20

15

20 4 6 8 10 12 14 16 18 20
Days

• • •
• • • •

•

•

••

Violent Cartoons
No Cartoons

FIGURE 3-15 This alternating-treatments design shows the frequency of aggressive behavior on days when chil-
dren watched violent cartoons compared with days when they did not watch cartoons. The level of
the aggressive behavior is greater on days with violent cartoons than on days with no cartoons.

Graphing Behavior and Measuring Change 55

more. The boys rode exercise bikes and received points for the amount of pedal-
ing that they did on the bikes. They later exchanged the points for toys and other
rewards. In this study, each time the exercise performance criterion was raised (the
boys had to pedal more to earn points), the boys’ exercise level increased accord-
ingly, thus demonstrating the effect of treatment.

Summary of Research Designs
A-B One baseline and one treatment phase. Not a true research design

A-B-A-B Two (or more) baseline phases and two (or more) treatment phases for the
same behavior of one subject. Also called a reversal design

Multiple-
baseline-across-behaviors

Baseline and treatment phases for two or more different behaviors of one
subject. Treatment is staggered across behaviors.

Multiple-
baseline-across-subjects

Baseline and treatment phases for the same behavior of two or more
subjects. Treatment is staggered across subjects.

Multiple-
baseline-across-settings

Baseline and treatment phases for the same behavior of the same subject
in two or more settings. Treatment is staggered across settings.

Days

To
ta

l C
af

fe
in

e
In

ta
ke

(m
g)

200

300

400

500

600

700

800

900

1000

1100

1200

1300

1 5 10 15 20 25 30 35 40 44 58 72 86 100 114 128 310 324

Baseline Treatment Phases Follow-Up

••• •

•
•

•

•

•

•

•

• •••

•

••
•
•

•

•

•
••••

•
•• •••

••

••

•••• • • • • •

•
• • • • •

1 2 3 4

Subject 1Subject 1

FIGURE 3-16 This changing-criterion graph shows that caffeine consumption decreased to a level below the cri-
terion each time the criterion was reduced. The solid horizontal bars in treatment phases 1-4 are
the criterion lines. The dashed lines show the mean level of the behavior in each phase. (From
Foxx, R. M., & Rubinoff, A. [1979]. Behavioral treatment of caffeinism: Reducing excessive coffee
drinking. Journal of Applied Behavior Analysis, 12, 335–344. Copyright © 1979 University of
Kansas Press. Reprinted by permission of Richard Foxx and the publisher.)

56 Chapter 3

Alternating-treatments
design

Baseline and treatment sessions are alternated rapidly. Baseline and
treatment sessions may occur on alternating days or may occur in dif-
ferent sessions on the same day.

Changing-criterion design A baseline phase and treatment phase for one subject. In the treatment
phase, there are progressive performance criteria or increasing goal levels
of the behavior.

CHAPTER SUMMARY

1. The six essential features of a complete behavior
modification graph are the y-axis and x-axis, labels
for the y-axis and x-axis, units for the y-axis
and x-axis, data points, phase lines, and phase
labels.

2. To graph behavioral data, you plot the data points
on the graph to reflect the level of the behavior on
the vertical axis (y-axis) and the unit of time on
the horizontal axis (x-axis).

3. The different dimensions of behavior you can
show on a graph include the frequency, duration,
intensity, and latency of the behavior. A graph
may also show the percentage of intervals of the
behavior derived from interval recording or time
sample recording or the percentage of opportu-
nities in which the behavior occurred (e.g., per-
centage correct).

4. A functional relationship between the treatment
(independent variable) and the target behavior
(dependent variable) exists when the treatment
causes the behavior to change. A functional
relationship or experimental control is demon-
strated when a target behavior changes after the
implementation of treatment and the treatment
procedure is repeated or replicated one or more
times and the behavior changes each time.

5. The different research designs you can use in behav-
ior modification research include the following:

â–  The A-B design shows baseline and treatment
for the behavior of one subject.

â–  The A-B-A-B design shows two baseline and
treatment phases repeated for the behavior of
one subject.

â–  A multiple-baseline design presents baseline and
treatment phases for one of the following
options: multiple behaviors of one subject, one
behavior of multiple subjects, or one behavior of
one subject across multiple settings. In each type
of multiple-baseline design, treatment is stag-
gered across behaviors, subjects, or settings.

â–  The alternating-treatments design presents data
from two experimental conditions that are rap-
idly alternated (baseline and treatment or two
treatments).

â–  Finally, in the changing-criterion design, a
baseline phase is followed by a treatment
phase in which sequential performance crite-
ria are specified.
All research designs, except the A-B design, con-

trol for the influence of extraneous variables, so that
the effectiveness of a treatment can be evaluated.

KEY TERMS

A-B design, 47
A-B-A-B reversal design, 48
abscissa, 40
alternating-treatments design

(ATD), 53
baseline, 42
changing-criterion design, 55

functional relationship, 47
graph, 39
multiple-baseline-across-

behaviors design, 49
multiple-baseline-across-settings

design, 49

multiple-baseline-across-subjects
design, 49

ordinate, 40
research design, 47

Graphing Behavior and Measuring Change 57

PRACTICE TEST

1. Why are graphs used in behavior modification
to evaluate behavior change? (p. 39)

2. What two variables are illustrated in a behavior
modification graph? (p. 40)

3. What is the y-axis? What is the x-axis? (p. 40)
4. What is labeled on the y-axis? On the x-axis?

(p. 40)
5. What is a phase? (p. 42)
6. Why are data points not connected across phase

lines? (p. 42)
7. Draw a hypothetical graph that illustrates the six

essential components of a behavior modification
graph. Label all six components on this hypotheti-
cal graph. (pp. 40–43)

8. What will you label the y-axis of a graph based
on interval recording? (p. 46)

9. What is an A-B design? What do A and B refer
to? (p. 47)

10. What is an A-B-A-B reversal design? Draw a
hypothetical A-B-A-B graph. Be sure all six com-
ponents are included. (p. 48)

11. What is a multiple-baseline design? Identify
three types of multiple-baseline designs. Draw a

hypothetical graph of a multiple-baseline-across-
subjects design. Be sure to include all six essen-
tial components. (pp. 49–51)

12. What is an extraneous variable? How does an
A-B-A-B design help you rule out extraneous
variables as the cause of the behavior change?
(pp. 47–49)

13. What does it mean to say that treatment is stag-
gered in a multiple-baseline design? (p. 50)

14. What is an alternating-treatments design (ATD)?
Draw a hypothetical graph of an ATD. Be sure to
include all six essential components. (p. 53)

15. How do you judge the effectiveness of treatment
in an ATD? (pp. 53–54)

16. Describe the changing-criterion design. Draw a
hypothetical graph of a changing-criterion
design. Include all six components. (pp. 55–56)

17. How do you determine that treatment is effec-
tive in a changing-criterion design? (p. 55)

18. What is a functional relationship? How do you
determine that a functional relationship exists
between a target behavior and a treatment pro-
cedure? (p. 47)

APPLICATIONS

1. In the application exercise in Chapter 2, you
developed a self-monitoring plan as the first
step in your self-management program. Once
you start to record your own target behavior,
the next step is to develop a graph and plot
your behavior on the graph each day. Some
people prefer to use a computer program to
generate a graph, but all that is really necessary
is a sheet of graph paper, a ruler, and a pencil.
Using a sheet of graph paper, prepare the graph
that you will use to plot the target behavior
from your self-management project. As you
develop your graph, be sure to observe the fol-
lowing rules.
a. Label the y-axis and x-axis appropriately.
b. Put the appropriate numbers on the y-axis

and the x-axis.

c. Ensure that the time period on the x-axis
covers at least 3 or 4 months so that you can
record the behavior for an extended period.

d. Plot the behavior on your graph every day as
you record the behavior.

e. Continue the baseline period for at least a
couple weeks so that any reactivity of the self-
monitoring stabilizes.

2. The data summary table in Figure 3-17 shows
the monthly total kilowatts of electricity used
by a fraternity house. In the two baseline phases,
no intervention was in place. In the two inter-
vention phases, the fraternity president gave
daily reminders to the fraternity brothers at
breakfast to turn out lights and turn off appli-
ances. Develop a graph from the data summary
table to show the effects of daily reminders on
the kilowatts of electricity used each month.

58 Chapter 3

3. Winifred worked with two autistic children who
engaged in self-injurious behavior (SIB) involv-
ing head-slapping. She recorded the frequency
of the SIB during baseline for both children,
Kale and Bud, and then implemented a treat-
ment involving reinforcement of alternative
behavior (see Chapter 15) and continued to col-
lect data for a period of time. The frequency of
SIB for Kale was 25, 22, 19, 19, 22, 22, and 23

in baseline and 12, 10, 5, 6, 5, 2, 1, 1, 1, 1, 0, 0,
1, 1, 0, 0, 0, and 0 during treatment. The fre-
quency of SIB for Bud was 12, 12, 15, 14, 13,
12, 12, 13, 10, 12, 14, and 17 in baseline and 5,
3, 4, 2, 0, 2, 0, 0, 0, 2, 0, 0, and 0 during treat-
ment. Draw the graph of the SIB data for Kale
and Bud. What kind of research design did
Winifred use when she provided treatment for
the SIB?

MISAPPLICATIONS

1. The Acme Widget Company was near bank-
ruptcy. Ace Consultants were called in to help.
They collected baseline data on employee pro-
ductivity for 4 weeks and determined that the
employees were assembling widgets only half
as fast as they were able to work. They imple-
mented an incentive system, and employee pro-
ductivity doubled. After 8 weeks of doubled
productivity, the Acme Company was making a
profit again. Ace Consultants decided to take
away the incentive system and return to baseline
for 4 weeks, and then re-implement the incen-
tive system (A-B-A-B research design) so that
they could determine whether the incentive
system caused the increase in productivity

or whether some extraneous variable was
responsible.
a. What is wrong with the use of an A-B-A-B

research design in this case?
b. What would you do if you worked for Ace

Consultants?
2. Alice was starting a self-management project to

increase the amount of running she did each
week. She planned to record her behavior for 2
or 3 weeks as a baseline before she implemented
an intervention. She decided that she would keep
a log of the distance that she ran every day and
plot her running distance on a graph each week.
She kept the log on her desk and wrote down the
duration of her run immediately after she ran. She

Months

Kilowatts
(rounded
to nearest
100)

1

4100

2

3900

3

4100

4

4200

5

3100

6

3000

7

2900

8

3000

9

2900

Baseline Intervention

Months

Kilowatts
(rounded
to nearest
100)

10

3800

11

3900

12

3800

13

2900

14

2900

15

2800

16

2900

Baseline Intervention

FIGURE 3-17 Data summary table showing kilowatts of electricity used per month across two baseline and two
intervention phases.

Graphing Behavior and Measuring Change 59

put her graph on the door to her room, and at the
end of each week, on Sunday night, she plotted
the number of miles she had run for the last
7 days. What was Alice doing wrong?

3. Dr. Pete was investigating an intervention for
improving social skills in socially anxious college
students. He identified three important types of
social behavior that he wanted to increase in
his subjects: initiating conversations, answering
questions, and smiling. He decided to use a

multiple-baseline-across-behaviors design in his
experiment. He would record all three behaviors
in each subject in a baseline before intervention.
He would then implement the intervention for all
three behaviors at one time and continue to
record the behaviors to see whether they increased
after the intervention was implemented.
a. What mistake did Dr. Pete make in his

multiple-baseline design?
b. What should he do differently?

60 Chapter 3

Chapter Four

Reinforcement

This chapter focuses on the basic behavioral principle of reinforcement. Scien-tific research has established a number of basic principles that explain the
behavior of people and other animals. Reinforcement is one of the first basic prin-
ciples that were systematically investigated by behavioral scientists, and it is a com-
ponent of many applications of behavior modification described in this text.
Reinforcement is the process in which a behavior is strengthened by the immedi-
ate consequence that reliably follows its occurrence. When a behavior is strength-
ened, it is more likely to occur again in the future.

Perhaps the earliest demonstration of reinforcement was reported by Thorndike
in 1911. Thorndike placed a hungry cat in a cage and put food outside of the cage

where the cat could see it. Thorndike rigged the cage so that a door
would open if the cat hit a lever with its paw. The cat was clawing
and biting the bars of the cage, reaching its paws through the open-
ings between the bars, and trying to squeeze through the opening.
Eventually, the cat accidentally hit the lever, the door opened, and
the cat got out of the cage and ate the food. Each time Thorndike
put the hungry cat inside the cage it took less time for the cat to hit
the lever that opened the door. Eventually, the cat hit the lever with
its paw as soon as Thorndike put it in the cage (Thorndike, 1911).
Thorndike called this phenomenon the law of effect.

In this example, when the hungry cat was put back in the cage
(Figure 4-1), the cat was more likely to hit the lever because this
behavior had resulted in an immediate consequence: escaping the

cage and getting food. Getting to the food was the consequence that reinforced
(strengthened) the cat’s behavior of hitting the lever with a paw.

â–  What is the principle of reinforcement?

â–  How is positive reinforcement different
from negative reinforcement?

â–  How are unconditioned reinforcers
different from conditioned reinforcers?

â–  What factors influence the
effectiveness of reinforcement?

â–  What are intermittent schedules of
reinforcement, and how do they affect
the rate of behavior?

Response

Reinforcement

Outcome: Behavior is more likely to occur in the future.

61

Starting in the 1930s, B. F. Skinner conducted numerous studies on the principle
of reinforcement in laboratory animals such as rats and pigeons (Skinner, 1938, 1956).
For example, in experiments with rats, Skinner placed the animal in an experimental
chamber and delivered a pellet of food each time the rat pressed a lever located on one
of the walls of the chamber. At first, the rat explored the box by moving around,
sniffing, climbing up on its hind legs, and so on. When it happened to press the lever
with one of its paws, the device automatically delivered a pellet of food to an opening
in the wall. Each time the hungry rat pressed the lever, it received a pellet of food.
Thus, the rat was more likely to press the lever each time it was placed in the chamber.
This one behavior, pressing the lever, was strengthened because when it occurred, it
was immediately followed by the receipt of food. The behavior of pressing the lever
increased in frequency relative to all the other behaviors the rat had exhibited when
put in the chamber.

FIGURE 4-1 A hungry cat is in the cage with food outside. When the cat hits the lever, the cage door opens and
the cat eats the food. As a result, the cat is more likely to hit the lever when it is put into the cage.

Response

The cat hits the lever with a paw and immediately the door opens and food is available.

Outcome: The cat is more likely to hit the lever when it is put in the cage in the future.

Response

The rat presses the lever and immediately food is presented.

Outcome: The rat is more likely to press the lever in the future.

62 Chapter 4

Defining Reinforcement

The examples of Thorndike’s cat and Skinner’s rat illustrate clearly the principle
of reinforcement. When a behavior results in a favorable outcome (one that con-
tributes to the well-being or survival of the animal), that behavior is more likely
to be repeated in the future in similar circumstances. Although the principle of
reinforcement was first systematically illustrated in laboratory animals, reinforce-
ment is a natural process that also influences human behavior. In Science and
Human Behavior (1953a), Skinner discussed the role of reinforcement in deter-
mining a wide variety of human behaviors. As stated by Sulzer-Azaroff and Mayer
(1991), reinforcement may occur naturally, as a result of our day-to-day interac-
tions with our social and physical environment, or it may be planned as part of
a behavior modification program used to change a person’s behavior. Table 4-1
provides examples of reinforcement.

As you can see from each of the examples in Table 4-1, reinforcement is
defined as follows:

1. The occurrence of a particular behavior
2. is followed by an immediate consequence
3. that results in the strengthening of the behavior. (The person is more likely

to engage in the behavior again in the future.)

We can determine that a behavior is strengthened when there is an increase
in its frequency, duration, intensity, or speed (decreased latency). A behavior that
is strengthened through the process of reinforcement is called an operant behav-
ior. An operant behavior acts on the environment to produce a consequence and,
in turn, is controlled by, or occurs again in the future as a result of, its immediate
consequence. The consequence that strengthens an operant behavior is called a
reinforcer.

In the first example in Table 4-1, the child cried at night when her parents
put her to bed. The child’s crying was an operant behavior. The reinforcer for
her crying was the parents’ attention. Because crying at night resulted in this
immediate consequence (reinforcer), the child’s crying was strengthened: She
was more likely to cry at night in the future.

For each of the other examples in Table 4-1, identify the operant behavior and the
reinforcer. The answers are available in Appendix A at the end of this chapter.

The graph in Figure 4-2 presents hypothetical data showing the effect of rein-
forcement on behavior. Notice that the frequency of the behavior is low during
baseline and higher during the reinforcement phase. As illustrated in Figure 4-2,
when the occurrence of a behavior is reinforced, it increases in frequency over
time. Other dimensions of a behavior (duration, intensity, speed) may also
increase as a function of reinforcement.

The graph in Figure 4-3 shows the effect of reinforcement on the duration
of a behavior. This graph, from a study by Liberman, Teigen, Patterson, and Baker
(1973), shows the duration of rational (nondelusional) talk by patients with schizo-
phrenia who were being treated in an institution. Liberman and colleagues mea-
sured the duration of rational talk during conversations with nurses. Liberman

Reinforcement 63

Days

Fr
eq

ue
nc

y
of

B
eh

av
io

r

0

2

4

6

8

10

0 105 15

Baseline Treatment

• •

• • •

• •
•

• •
• • • •

•

FIGURE 4-2 This graph with hypothetical data shows the effect of reinforcement on the frequency of a behavior.
When reinforcement is used after a baseline phase, the behavior increases in frequency.

TABLE 4-1 Examples for Self-Assessment (Reinforcement)

1. A child cries at night after being put to bed and her parents come to her room to comfort her and
calm her down. As a result, the child now cries more often at bedtime.

2. A woman waiting for a bus opens up her umbrella when it rains. The umbrella keeps the rain from
hitting her. Now she always opens up her umbrella when it rains.

3. When a chef cooks well-done steaks, it creates smoke. He turns on the exhaust fan, and the smoke
is sucked out of the kitchen. He is now more likely to turn on the fan when he cooks well-done
steaks.

4. A college student is answering study guide questions for her behavior modification class. When she
can’t figure out an answer to a question, she asks her friend who already took the class. Her friend
tells her the correct answer. As a result, she is more likely to ask her friend for answers to questions
she doesn’t know.

5. A teacher smiles at Johnny and praises him when he stays in his seat and pays attention in the
classroom. As a result, Johnny is more likely to sit in his seat and pay attention (i.e., to look at his
teacher when she teaches).

6. When Patricia goes to bed while her husband watches TV, the noise keeps her up. She uses ear-
plugs to eliminate the TV noise and is able to fall asleep. As a result, she is more likely to use ear-
plugs when she goes to bed with the TV on.

7. Instead of paying workers by the hour, a bicycle manufacturing company begins paying piece rate, in
which workers on the assembly line earn a certain amount of money for each bicycle they assemble.
As a result, the workers assemble more bicycles each day and earn more money.

8. A 2-year-old child has a tantrum (crying and screaming) in the grocery store when he demands
candy and his mother says no. His mother eventually buys him the candy and he stops his tantrum.
As a result, the mother is more likely to give him candy when he demands it and has a tantrum. In
addition, the child is more likely to have a tantrum in the store because it results in receiving candy
from his mother.

64 Chapter 4

wanted to reinforce rational talk so that it would increase, and the schizophrenic
patients would thus appear more normal. In this study, rational talk was reinforced
by the nurses with attention and one-on-one chats during snack time. At the same
time, delusional talk was not reinforced (the nurses withheld the social attention).
Figure 4-3 shows that rational talk increased in duration during the treatment
phase, when social reinforcement was used.

•
••••

•

••

••
•

• •• •

•

• •
• •

•
••

•

•
•

•

• • •
•

••••••••
•• ••

•
•

•
•

•
•

• •

•

• •

•
•

•
•

•
•

••

••

•
•

•
•

•
•

•
•

•

• ••

•
•

• •
• •

•
• • •

••
•

• • •
•

•
••

•
••

•
•

••

•

•
•

•
•

•
• •

•
• •

•

•• •
•

•
•

•
•

•

M
in

ut
es

o
f R

at
io

na
l T

al
k

pe
r D

ay

Baseline Treatment

Jane

Herman

Jack

Number of Days

0

10

20

30

40

2 4 6 8 10 12 14 16 18 20 22 24 26 28 30

Mary

0

10

20

30

40

0

10

20

30

40

0

10

20

30

40

FIGURE 4-3 This graph of a multiple-baseline-across-subjects design shows the effect of reinforcement on
the duration of rational talk by four patients with schizophrenia. Note that the duration of rational
talk increased for all four subjects when reinforcement was used (treatment). (From Liberman,
R. P., Teigen, J., Patterson, R., & Baker, V. [1973]. Reducing delusional speech in chronic paranoid
schizophrenics. Journal of Applied Behavior Analysis, 6, 57–64. Copyright © 1973 University
of Kansas Press. Reprinted by permission of Robert P. Liberman, James R. Teigen, and the
publisher.)

Reinforcement 65

What type of research design is illustrated in the graph in Figure 4-3?

Figure 4-3 presents a multiple-baseline-across-subjects design. There is a base-
line and treatment (reinforcement) phase for each of four patients. The implemen-
tation of the reinforcement procedure is staggered over time for the four patients.

On Terms: Reinforce Behavior, not People
â–  It is correct to say that you reinforce a behavior (or a response). You are strengthening a behavior

by reinforcing it. To say “The teacher reinforced standing in line quietly with praise” is correct.
■ It is incorrect to say that you reinforce a person. You don’t strengthen a person; you strengthen a

person’s behavior. To say, “The teacher reinforced Sarah for standing in line quietly” is not correct.

Now that you understand the basic definition of reinforcement, it is important
to understand the distinction between positive reinforcement and negative
reinforcement.

Positive and Negative Reinforcement

There are two types of reinforcement: positive reinforcement and negative rein-
forcement. It is extremely important to remember that both positive reinforcement
and negative reinforcement are processes that strengthen a behavior; that is, they
both increase the probability that the behavior will occur in the future. Positive
and negative reinforcement are distinguished only by the nature of the conse-
quence that follows the behavior.

Positive reinforcement is defined as follows.

1. The occurrence of a behavior
2. is followed by the addition of a stimulus (a reinforcer) or an increase in

the intensity of a stimulus,
3. which results in the strengthening of the behavior.

Negative reinforcement, by contrast, is defined as follows.

1. The occurrence of a behavior
2. is followed by the removal of a stimulus (an aversive stimulus) or a

decrease in the intensity of a stimulus,
3. which results in the strengthening of the behavior.

A stimulus is an object or event that can be detected by one of the senses,
and thus has the potential to influence the person (stimuli is the plural form of
the word stimulus). The object or event may be a feature of the physical environ-
ment or the social environment (the behavior of the person or of others).

In positive reinforcement, the stimulus that is presented or that appears after
the behavior is called a positive reinforcer. (A positive reinforcer often is seen as
something pleasant, desirable, or valuable that a person will try to get.) In negative
reinforcement, the stimulus that is removed or avoided after the behavior is called
an aversive stimulus. (An aversive stimulus often is seen as something unpleasant,

66 Chapter 4

painful, or annoying that a person will try to get away from or avoid.) The essen-
tial difference, therefore, is that in positive reinforcement, a response produces a
stimulus (a positive reinforcer), whereas in negative reinforcement, a response
removes or prevents the occurrence of a stimulus (an aversive stimulus). In both
cases, the behavior is more likely to occur in the future.

Consider Example 8 in Table 4-1. The mother’s behavior of buying her child
candy results in termination of the child’s tantrum (an aversive stimulus is
removed). As a result, the mother is more likely to buy her child candy when he
tantrums in a store. This is an example of negative reinforcement. On the other
hand, when the child tantrums, he gets candy (a positive reinforcer is presented).
As a result, he is more likely to tantrum in the store. This is an example of positive
reinforcement.

Some people confuse negative reinforcement and punishment (see Chapter 6).
They are not the same. Negative reinforcement (like positive reinforcement)
increases or strengthens a behavior. Punishment, in contrast, decreases or weakens
a behavior. The confusion comes from the use of the word negative in negative
reinforcement. In this context, the word negative does not mean bad or unpleasant,
but simply refers to the removal (subtraction) of the stimulus after the behavior.

Numerous examples of positive and negative reinforcement abound in our
everyday lives. Of the eight examples in Table 4-1, five illustrate positive reinforce-
ment and four illustrate negative reinforcement (example 8 illustrates both).

Read each example in Table 4-1. Which ones are examples of positive reinforce-
ment? Which are examples of negative reinforcement? Explain your selections.
The answers may be found in Appendix B at the end of this chapter.

The important thing to remember about positive reinforcement and negative
reinforcement is that both have the same impact on the behavior: They strengthen
it. Reinforcement is always defined by the effect it has on the behavior (Skinner,
1958). This is called a functional definition. Consider the following example: A
child completes an academic task independently and his teacher walks up to his
desk, says “Good job,” and pats him on the back.

Is this scenario an example of positive reinforcement?

In this case, we cannot tell because not enough information is presented.
This situation would be an example of positive reinforcement only if, as a result
of the praise and pat on the back, the child was more likely to complete academic
tasks independently in the future. Remember, this is the functional definition of
reinforcement: The consequence of a behavior increases the probability that the
behavior will occur again in the future. For most children, praise and teacher
attention are reinforcers that would strengthen the behavior of completing aca-
demic tasks. However, for some children (some children with autism, for exam-
ple), teacher attention may not be a reinforcer. Therefore, praise and a pat on
the back would not strengthen the behavior (Durand, Crimmins, Caufield, &
Taylor, 1989). Durand and his colleagues illustrated that to determine whether a
particular consequence will be a reinforcer for a particular person, you have to
try it out and measure its effect on the behavior. Working with children who had
severe developmental disorders, they compared two consequences for their

Reinforcement 67

academic performance. Sometimes the children received praise for correct perfor-
mance, and sometimes correct performance resulted in a brief break from the aca-
demic task. Durand and colleagues found that praise increased correct
performance for some children but not for others, and that the brief break
(removal of the academic demand) increased correct performance for some chil-
dren but not for others. Durand emphasized the importance of identifying reinfor-
cers by measuring their effects on the behavior.

Whenever you have to analyze a situation and determine whether it illustrates
positive or negative reinforcement, ask yourself three questions:

1. What is the behavior?
2. What happened immediately after the behavior? (Was a stimulus added or

removed?)
3. What happened to the behavior in the future? (Was the behavior strength-

ened? Was it more likely to occur?)

If you can answer each of these three questions, you can identify an example
as positive reinforcement, negative reinforcement, or neither.

Social versus Automatic Reinforcement
As you have learned, reinforcement can involve the addition of a reinforcer (posi-
tive reinforcement) or the removal of an aversive stimulus (negative reinforce-
ment) following the behavior. In both cases, the behavior is strengthened. For
both positive and negative reinforcement, the behavior may produce a conse-
quence through the actions of another person or through direct contact with the
physical environment (e.g., Iwata, Vollmer, & Zarcone, 1990; Iwata, Vollmer,
Zarcone, & Rodgers, 1993). When a behavior produces a reinforcing conse-
quence through the actions of another person, the process is social reinforcement.
An example of social positive reinforcement might involve asking your roommate
to bring you the bag of chips. An example of social negative reinforcement might
involve asking your roommate to turn down the TV when it is too loud. In both
cases, the consequence of the behavior was produced through the actions of
another person. When the behavior produces a reinforcing consequence through
direct contact with the physical environment, the process is automatic reinforce-
ment. An example of automatic positive reinforcement would be if you went to
the kitchen and got the chips for yourself. An example of automatic negative rein-
forcement would be if you got the remote and turned down the volume on the
TV yourself. In both cases, the reinforcing consequence was not produced by
another person.

One type of positive reinforcement involves the opportunity to engage in a
high-probability behavior (a preferred behavior) as a consequence for a low-
probability behavior (a less-preferred behavior), to increase the low-probability
behavior (Mitchell & Stoffelmayr, 1973). This is called the Premack principle
(Premack, 1959). For example, the Premack principle operates when parents
require their fourth grade son to complete his homework before he can go outside
to play with his friends. The opportunity to play (a high-probability behavior) after
the completion of the homework (low-probability behavior) reinforces the

68 Chapter 4

behavior of doing homework; that is, it makes it more likely that the child will
complete his homework.

On Terms: Distinguishing Between Positive and Negative Reinforcement

Some students have confusion distinguishing between positive and negative reinforcement. They both are
types of reinforcement, therefore, they both strengthen behavior. The only difference is whether a stimu-
lus is added (positive reinforcement) or removed (negative reinforcement) following the behavior. Think of
positive as a plus or addition (þ) sign and negative as a minus or subtraction (�) sign. In þ reinforce-
ment, you add a stimulus (a reinforcer) after the behavior. In – reinforcement, you subtract or take away
a stimulus (an aversive stimulus) after the behavior. If you think of positive and negative in terms of add-
ing or subtracting a stimulus after the behavior, the distinction should be more clear.

Escape and Avoidance Behaviors

When defining negative reinforcement, a distinction is made between escape and
avoidance. In escape behavior, the occurrence of the behavior results in the ter-
mination of an aversive stimulus that was already present when the behavior
occurred. That is, the person escapes from the aversive stimulus by engaging in a
particular behavior, and that behavior is strengthened. In avoidance behavior, the
occurrence of the behavior prevents an aversive stimulus from occurring. That is,
the person avoids the aversive stimulus by engaging in a particular behavior, and
that behavior is strengthened.

In an avoidance situation, a warning stimulus often signals the occurrence of
an aversive stimulus, and the person engages in an avoidance behavior when this
warning stimulus is present. Both escape and avoidance are types of negative rein-
forcement; therefore, both result in an increase in the rate of the behavior that ter-
minated or avoided the aversive stimulus.

The distinction between escape and avoidance is shown in the following situ-
ation. A laboratory rat is placed in an experimental chamber that has two sides
separated by a barrier; the rat can jump over the barrier to get from one side to
the other. On the floor of the chamber is an electric grid that can be used to
deliver a shock to one side or the other. Whenever the shock is presented on the
right side of the chamber, the rat jumps to the left side, thus escaping from the
shock. Jumping to the left side of the chamber is escape behavior because the rat
escapes from an aversive stimulus (the shock). When the shock is applied to the
left side, the rat jumps to the right side. The rat learns this escape behavior rather
quickly and jumps to the other side of the chamber as soon as the shock is
applied.

In the avoidance situation, a tone is presented just before the shock is applied.
(Rats have better hearing than vision.)

What does the rat learn to do when the tone is presented?

After a number of instances in which the tone is presented just before the
shock, the rat starts to jump to the other side of the chamber as soon as it hears
the tone. The tone is the warning stimulus; the rat avoids the shock by jumping
to the other side as soon as the warning stimulus is presented.

Reinforcement 69

Everyday Examples of Escape and Avoidance Behaviors

Escape A person steps barefoot on the hot asphalt and immediately steps onto the grass.
Stepping onto the grass results in escape from the heat of the hot asphalt.

Avoidance A person puts on shoes the next time she walks on hot asphalt. Wearing shoes
results in avoidance of the heat from the hot asphalt.

Escape You start your car and the radio blasts on because someone left the volume all
the way up. You turn down the volume to escape the ear-piercing noise.

Avoidance You turn down the volume on the car radio before you start the car. In this case,
you avoid the noise from the radio.

Escape You sit down in a movie theater near a large group of kids. They are very loud
during the movie, so you move to a seat far away from them to escape the noise.

Avoidance You walk into a movie theater and take a seat far away from a group of kids. In
this way, you avoid the noise they make.

Conditioned and Unconditioned Reinforces

Reinforcement is a natural process that affects the behavior of humans and other
animals. Through the process of evolution, we have inherited certain biological
characteristics that contribute to our survival. One characteristic we have inherited
is the ability to learn new behaviors through reinforcement. In particular, certain
stimuli are naturally reinforcing because the ability of our behaviors to be rein-
forced by these stimuli has survival value (Cooper, Heron, & Heward, 1987;
2007). For example, food, water, and sexual stimulation are natural positive rein-
forcers because they contribute to survival of the individual and the species.
Escape from painful stimulation or extreme levels of stimulation (cold, heat, or
other discomforting or aversive stimulation) is naturally negatively reinforcing

Response

Outcome: The rat is more likely to jump to the other side in the future when a tone is presented.

When the tone is presented, the rat jumps to the other side and immediately the rat avoids the electric shock.

AvoidanceAvoidance

Response

Outcome: The rat is more likely to jump to the other side in the future when shock is presented.

When a shock is given, the rat jumps to the other side and immediately the rat escapes from the electric shock.

EscapeEscape

70 Chapter 4

because escape from or avoidance of these stimuli also contributes to survival.
These natural reinforcers are called unconditioned reinforcers because they func-
tion as reinforcers the first time they are presented to most human beings; no prior
experience with these stimuli is needed for them to function as reinforcers.
Unconditioned reinforcers sometimes are called primary reinforcers. These stimuli
are unconditioned reinforcers because they have biological importance (Cooper
et al., 1987; 2007).

Another class of reinforcers is the conditioned reinforcers. A conditioned
reinforcer (also called a secondary reinforcer) is a stimulus that was once neutral
(a neutral stimulus does not currently function as a reinforcer; i.e., it does not
influence the behavior that it follows) but became established as a reinforcer by
being paired with an unconditioned reinforcer or an already established condi-
tioned reinforcer. For example, a parent’s attention is a conditioned reinforcer for
most children because attention is paired with the delivery of food, warmth, and
other reinforcers many times in the course of a young child’s life. Money is per-
haps the most common conditioned reinforcer. Money is a conditioned reinforcer
because it can buy (is paired with) a wide variety of unconditioned and condi-
tioned reinforcers throughout a person’s life. If you could no longer use money
to buy anything, it would no longer be a conditioned reinforcer. People would
not work or engage in any behavior to get money if it could not be used to obtain
other reinforcers. This illustrates one important point about conditioned reinfor-
cers: They continue to be reinforcers only if they are at least occasionally paired
with other reinforcers.

Nearly any stimulus may become a conditioned reinforcer if it is paired with
an existing reinforcer. For example, when trainers teach dolphins to perform tricks
at aquatic parks, they use a handheld clicker to reinforce the dolphin’s behavior.
Early in the training process, the trainer delivers a fish as a reinforcer and pairs
the sound of the clicker with the delivery of the fish to eat. Eventually, the click-
ing sound itself becomes a conditioned reinforcer. After that, the trainer occasion-
ally pairs the sound with the unconditioned reinforcer (the fish) so that the
clicking sound continues to be a conditioned reinforcer (Pryor, 1985). A neutral
stimulus such as a plastic poker chip or a small square piece of colored cardboard
can be used as a conditioned reinforcer (or token) to modify human behavior in a
token reinforcement program. In a token reinforcement program, the token is pre-
sented to the person after a desirable behavior, and later the person exchanges the
token for other reinforcers (called backup reinforcers). Because the tokens are
paired with (exchanged for) the backup reinforcers, the tokens themselves become
reinforcers for the desirable behavior. (See Kazdin [1982] for a review of research
on token reinforcement programs.) Chapter 22 explains token reinforcement pro-
grams in more detail.

When a conditioned reinforcer is paired with a wide variety of other reinfor-
cers, it is called a generalized conditioned reinforcer. Money is a generalized
conditioned reinforcer because it is paired with (exchanged for) an almost unlim-
ited variety of reinforcers. As a result, money is a powerful reinforcer that is less
likely to diminish in value (to become satiated) when it is accumulated. That is,
satiation (losing value as a reinforcer) is less likely to occur for generalized reinfor-
cers such as money. Tokens used in a token economy are another example of a

Reinforcement 71

generalized conditioned reinforcer because they are exchanged for various other
backup reinforcers. As a result, people can accumulate tokens without rapid satia-
tion. Praise is also a generalized conditioned reinforcer because praise is paired
with numerous other reinforcers across a person’s lifetime.

Factors That Influence the Effectiveness
of Reinforcement

The effectiveness of reinforcement is influenced by a number of factors. These
include the immediacy and consistency of the consequence, establishing opera-
tions, the magnitude of the reinforcer, and individual differences.

Immediacy
The time between the occurrence of a behavior and the reinforcing consequence
is important. For a consequence to be most effective as a reinforcer, it should
occur immediately after the response occurs. (One instance of a behavior is called
a response). The longer the delay between the response and the consequence, the
less effective the consequence will be because the contiguity or connection
between the two is weakened. If the time between the response and the conse-
quence becomes too long and there is no contiguity, the consequence will not
have an effect on the behavior. For example, if you wanted to teach your dog to
sit on command and you gave the dog a treat 5 minutes after it performed the
behavior, the treat would not function as a reinforcer for sitting. In this case, the
delay would be too long. Rather, the treat would function as a reinforcer for what-
ever behavior the dog engaged in immediately before receiving the treat (probably
begging, which is the behavior usually reinforced with treats). Conversely, if you
gave the dog a treat immediately after it sat, the treat would reinforce sitting
behavior, and the dog would be more likely to sit in the future when given the
corresponding command.

Consider the importance of immediate reinforcement on social behavior.
When you talk to someone, you receive immediate social responses from the lis-
tener, such as smiles, head nods, eye contact, and laughter, which reinforce saying
the things you say. These social reinforcers strengthen your appropriate social
behavior. You learn what is appropriate to say and what is inappropriate, accord-
ing to the immediate response of the listener. For example, if you tell a joke and
people laugh, you are more likely to repeat the joke in the future. If you don’t get
immediate laughs, you will be less likely to tell the joke in the future.

Contingency
If a response is consistently followed by an immediate consequence, that conse-
quence is more likely to reinforce the response. When the response produces the
consequence and the consequence does not occur unless the response occurs first,
we say that a contingency exists between the response and the consequence.
When a contingency exists, the consequence is more likely to reinforce the
response (e.g., see Borrero, Vollmer & Wright, 2002). Consider the example of

72 Chapter 4

turning the key in your ignition to start your car. This is an example of contin-
gency: Every time you turn the key, the car starts. The behavior of turning the
key is reinforced by the engine starting. If the engine started only sometimes
when you turned the key, and if it started sometimes when you did not turn the
key, the behavior of turning the key in this particular car would not be strength-
ened very much. A person is more likely to repeat a behavior when it results in a
consistent reinforcing consequence. That is, a behavior is strengthened when a
reinforcer is contingent on the behavior (when the reinforcer occurs only if the
behavior occurs).

Motivating Operations
Some events can make a particular consequence more or less reinforcing at some
times than at other times. These antecedent events, called motivating operations
(MOs), alter the value of a reinforcer. There are two types of MOs; establishing
operations and abolishing operations. An establishing operation (EO) makes a
reinforcer more potent (it establishes the effectiveness of a reinforcer). An abolish-
ing operation (AO) makes a reinforcer less potent (it abolishes or decreases the
effectiveness of a reinforcer). Motivating operations have two effects: a) they alter
the value of a reinforcer and b) they make the behavior that produces that rein-
forcer more or less likely to occur at that time. An EO makes a reinforcer more
potent and makes a behavior that produces the reinforcer more likely. An AO
makes a reinforcer less potent and makes a behavior that produces that reinforcer
less likely.

Let’s consider some examples of establishing operations. Food is a more pow-
erful reinforcer for a person who hasn’t eaten recently. Not having eaten in a
while is an EO that makes food more reinforcing at that time and makes the
behavior of getting food more likely to occur. Likewise, water is a more potent
reinforcer for someone who has not had a drink all day or who just ran 6 miles.
Water or other beverages are more reinforcing when a person just ate a large
amount of salty popcorn than when a person did not. (That is why some bars
give you free salty popcorn.) In these examples, going without food or water (dep-
rivation), running 6 miles, and eating salty popcorn are events called establishing
operations because they increase the effectiveness of a reinforcer at a particular
time or in a particular situation and make the behavior that results in that rein-
forcer more likely to occur.

Deprivation is a type of establishing operation that increases the effectiveness
of most unconditioned reinforcers and some conditioned reinforcers. A particular
reinforcer (such as food or water) is more powerful if a person has gone without
it for some time. For example, attention may be a more powerful reinforcer for a
child who has gone without attention for a period of time. Similarly, although
money is almost always a reinforcer, it may be a more powerful reinforcer for
someone who has gone without money (or enough money) for a period of time.
In addition, any circumstances in which a person needs more money (e.g., unex-
pected doctor bills) make money a stronger reinforcer.

Let’s consider some examples of abolishing operations. Food is not likely to
be reinforcing right after a person just finished a large meal. Having just eaten a
large meal is an AO that makes food less reinforcing at that time and makes the

Reinforcement 73

behavior of getting food less likely to occur. Water or other drinks are not likely to
be reinforcing for someone who has just drank a substantial amount of water.
Drinking a large amount of water makes water less reinforcing at that time and
makes the behavior of getting water less likely to occur. These events are called
abolishing operations because they abolish the effectiveness of a reinforcer at a
particular time or in a particular situation and make the behavior that results in
that reinforcer less likely to occur.

These examples illustrate a type of abolishing operation called satiation. Sati-
ation occurs when a person has recently consumed a large amount of a particular
reinforcer (such as food or water) or has had substantial exposure to a reinforcing
stimulus. As a result, these reinforcers are less potent at that time. For example,
your favorite music may be less reinforcing if you have listened to it for the past
5 hours. Likewise, adult attention may be less reinforcing to a child who has just
received substantial one-on-one attention from a teacher. Although substantial
exposure to, or consumption of, a reinforcer decreases the effectiveness of a rein-
forcer, the effects of satiation diminish over time. The longer it has been since
the reinforcer was consumed, the more powerful the reinforcer becomes.

Instructions or rules may also function as an establishing operation or abolish-
ing operation and influence the reinforcing value of a stimulus (Schlinger, 1993).
For example, pennies are not potent reinforcers for most people. However, if you
were told that there was a copper shortage and that pennies were now worth
50 cents apiece, the reinforcing value of pennies would increase, and you would
be more likely to engage in behavior that resulted in obtaining more pennies.

Is the preceding scenario an example of an establishing operation or an abolish-
ing operation?

This scenario represents an EO because the reinforcing value of pennies was
increased.

Consider another example. Suppose that your friend had some tickets for
events at an amusement park you were about to attend. If you were told that the
tickets had expired and were no longer being accepted, the reinforcing value of the
tickets would be lost and you would be less likely to ask your friend for the tickets.

Is the receding scenario an example of an establishing operation or an abolish-
ing operation?

This scenario represents an AO because the reinforcing value of the tickets
was decreased.

Consider another example. You have just bought a new table for your com-
puter and printer. When you read the assembly instructions and discover that you
need a screwdriver to assemble it, this increases the value of a screwdriver as a
reinforcer at that time. As a result, you are more likely to go look for a screwdriver.
Searching for a screwdriver is strengthened by finding it and successfully assem-
bling the table.

Is the preceding scenario an example of an establishing operation or an abolish-
ing operation?

This scenario represents an EO because the reinforcing value of a screwdriver
was increased.

74 Chapter 4

Establishing operations and abolishing operations also influence the effective-
ness of negative reinforcement. When an event increases the aversiveness of a
stimulus, escape from or removal of the stimulus becomes more reinforcing
(EO). When an event decreases the aversiveness of a stimulus, escape from or
removal of the stimulus becomes less reinforcing (AO). For example, a headache
may be an establishing operation that makes loud music more aversive; therefore,
turning off the loud music is more reinforcing when you have a headache. (You
are more likely to turn off loud music when you have a headache.) However,
being with friends on the weekend (and not having a headache) decreases the
aversiveness of loud music and makes turning off the loud music less reinforcing.
Consider another example. Sunshine probably is not aversive for most people, but
when a person has a bad sunburn, escape from the heat of the sun is more rein-
forcing. Therefore, the bad sunburn is an establishing operation that makes stay-
ing indoors or sitting in the shade more reinforcing because these behaviors
terminate the heat of the sun (aversive stimulus). On the other hand, applying
sunscreen may be an abolishing operation that decreases the aversiveness of being
in the sunshine and makes escape from the sun less reinforcing. (See Michael
[1982, 1993b] and Laraway, Snycerski, Michael, Poling [2003] for a more com-
plete discussion of establishing operations and abolishing operations.)

FOR FURTHER READING
Motivating Operations

Establishing operation (EO) is a term used to describe any event that makes a reinforcer more potent. An
EO also makes a behavior that produces that reinforcer more likely to occur. The concept of the EO was
described in detail in 1982 in an article by Jack Michael, and it has been written about extensively since
then (e.g., McGill, 1999). In his 1982 article, Michael defines the term EO and helps distinguish it from
another type of antecedent event, a discriminative stimulus, or SD (see Chapter 7). In a more recent
article, Laraway and colleagues (Laraway, Snycerski, Michael, Poling, 2003) refined the concept of EO
and discussed it in the larger context of motivation. They also introduced the concept of the abolishing
operation (AO). EOs and AOs are important because they are major influences on the effectiveness of
reinforcement and because EOs and AOs are often manipulated in behavior modification procedures to
help people change their behavior (see Chapters 13 and 16).

Individual Differences
The likelihood of a consequence being a reinforcer varies from person to person, so it
is important to determine that a particular consequence is a reinforcer for a particular
person. It is important not to assume that a particular stimulus will be a reinforcer for
a person just because it appears to be a reinforcer for most people. For example,
praise may be meaningless to some people, even though it is a reinforcer for most.
Chocolate candy may be reinforcers for most children, but it won’t be for the child
who is allergic to chocolate and gets sick when she eats it. Chapter 15 discusses vari-
ous ways to identify which consequences function as reinforcers for people.

Magnitude
The other characteristic of a stimulus that is related to its power as a reinforcer is
its amount or magnitude. Given the appropriate establishing operation, generally,

Reinforcement 75

the effectiveness of a stimulus as a reinforcer is greater if the amount or magnitude
of a stimulus is greater. This is true for both positive and negative reinforcement. A
larger positive reinforcer strengthens the behavior that produces it to a greater extent
than a smaller amount or magnitude of the same reinforcer does. For example, a
person would work longer and harder for a large amount of money than for a
small amount. Likewise, the termination of a more intense aversive stimulus
strengthens the behavior that terminates it more than the termination of a lower
magnitude or intensity of the same stimulus would. For example, a person would
work harder or engage in more behavior to decrease or eliminate an extremely pain-
ful stimulus than a mildly painful stimulus. You would work a lot harder to escape
from a burning building than you would to get out of the hot sun.

Factors That Influence the Effectiveness of Reinforcement

Immediacy A stimulus is more effective as a reinforcer when it is delivered immediately after
the behavior.

Contingency A stimulus is more effective as a reinforcer when it is delivered contingent on the
behavior.

Motivating operations Establishing operations make a stimulus more effective as a reinforcer at a par-
ticular time. Abolishing operations make a stimulus less potent as a reinforcer at
a particular time.

Individual differences Reinforcers vary from person to person.

Magnitude Generally, a more intense stimulus is a more effective reinforcer.

Schedules of Reinforcement

The schedule of reinforcement for a particular behavior specifies whether every
response is followed by a reinforcer or whether only some responses are followed
by a reinforcer. A continuous reinforcement schedule (CRF schedule) is one in
which each occurrence of a response is reinforced. In an intermittent reinforce-
ment schedule, by contrast, each occurrence of the response is not reinforced.
Rather, responses are occasionally or intermittently reinforced. Consider the fol-
lowing example. Maria was recently hired by a company that makes furniture,
and her job involves screwing knobs on cabinet doors. The first day on the job,
the supervisor showed Maria how to properly screw on the knobs. The supervisor
then watched Maria do the job for the first few minutes and praised her each time
she correctly screwed a knob on a cabinet door. This is a CRF schedule because
every response (screwing on a knob correctly) was followed by the reinforcing con-
sequence (praise from the supervisor). After Maria’s first few minutes on the job,
the supervisor left and came back occasionally during the day, watched Maria do
her job, and praised her when she screwed on a knob correctly. This is an inter-
mittent reinforcement schedule because Maria’s behavior of putting knobs on cab-
inet doors was reinforced only some of the times that it occurred.

In this example, you can see that a CRF schedule was used initially when
Maria was first learning the behavior. After Maria had learned the behavior (as

76 Chapter 4

determined by the fact that she performed it correctly each time), the supervisor
shifted to an intermittent reinforcement schedule. This illustrates the two different
uses of CRF and intermittent reinforcement schedules. A CRF schedule is used
when a person is learning a behavior or engaging in the behavior for the first
time. This is called acquisition: The person is acquiring a new behavior. Once
the person has acquired or learned the behavior, an intermittent reinforcement
schedule is used so that the person continues to engage in the behavior. This is
called maintenance: The behavior is maintained over time with the use of inter-
mittent reinforcement. A supervisor could not stand by Maria and praise her for
every correct behavior every day that she works. Not only is this impossible, but it
is also unnecessary. Intermittent reinforcement is more effective than a CRF
schedule for maintaining a behavior.

Describe how a vending machine illustrates a CRF schedule and a slot machine
illustrates an intermittent reinforcement schedule.

The behavior of putting money in a vending machine and pushing the selec-
tion button is reinforced every time it occurs because the machine gives you the
item that you paid for and selected. The behavior of putting money in a slot
machine and pulling the handle is reinforced on an intermittent schedule
because the slot machine pays off only occasionally (Figure 4-4).

Ferster and Skinner (1957) studied various types of intermittent reinforcement
schedules. In their experiments, pigeons in experimental chambers pecked round
disks (or keys) mounted on the wall of the chamber in front of them. The key

FIGURE 4-4 The slot machine works on an intermittent schedule of reinforcement. You do not hit the jackpot
and get money from the machine every time you put money in the machine. The vending machine
works on a continuous schedule of reinforcement. Every time you put money in the machine, you
get an item from the machine.

Reinforcement 77

could be illuminated, and the apparatus automatically recorded each key peck. As
reinforcers for key-pecking behavior, small amounts of food were delivered
through an opening in the wall below the key. Ferster and Skinner described
four basic types of schedules: fixed ratio, variable ratio, fixed interval, and variable
interval. Although these reinforcement schedules originally were studied with lab-
oratory animals, they are also applied to human behavior.

Fixed Ratio
In fixed ratio and variable ratio schedules of reinforcement, the delivery of the rein-
forcer is based on the number of responses that occur. In a fixed ratio (FR) sched-
ule, a specific or fixed number of responses must occur before the reinforcer is
delivered. That is, a reinforcer is delivered after a certain number of responses. For
example, in a fixed ratio 5 (FR 5) schedule, the reinforcer follows every fifth
response. In an FR schedule, the number of responses needed before the reinforcer
is delivered does not change. Ferster and Skinner (1957) found that pigeons would
engage in high rates of responding on FR schedules, however, there was often a
brief pause in responding after the delivery of the reinforcer. Ferster and Skinner
investigated FR schedules ranging from FR 2 to FR 400, in which 400 responses
had to occur before the reinforcer was delivered. Typically, the rate of responding is
greater when more responses are needed for reinforcement in an FR schedule.

FR schedules of reinforcement sometimes are used in academic or work set-
tings to maintain appropriate behavior. Consider the example of Paul, a 26-year-old
adult with severe intellectual disability who works in a factory packaging parts for
shipment. As the parts come by on a conveyor belt, Paul picks them up and puts
them into boxes. Paul’s supervisor delivers a token (conditioned reinforcer) after
every 20 parts that Paul packages. This is an example of an FR 20. At lunch and
after work, Paul exchanges his tokens for backup reinforcers (e.g., snacks or soft
drinks). An FR schedule could be used in a school setting by giving students rein-
forcers (such as stars, stickers, or good marks) for correctly completing a fixed num-
ber of problems or other academic tasks. Piece-rate pay in a factory, in which
workers get paid a specified amount of money for a fixed number of responses
(e.g., $5 for every 12 parts assembled), is also an example of an FR schedule.

Variable Ratio
In a variable ratio (VR) schedule, as in an FR schedule, delivery of a reinforcer is
based on the number of responses that occur, but in this case, the number of
responses needed for reinforcement varies each time, around an average number.
That is, a reinforcer is delivered after an average of x responses. For example, in a
variable ratio 10 (VR 10) schedule, the reinforcer is provided after an average of
10 responses. The number of responses needed for each reinforcer may range
from just 2 or 3 up to 20 or 25, but the average number of responses equals 10.
Ferster and Skinner (1957) evaluated VR schedules with pigeons and found that
such schedules produced high, steady rates of responding; in contrast with FR
schedules, there is little pausing after the delivery of the reinforcer. In their
research, Ferster and Skinner evaluated various VR schedules, including some
that needed a large number of responses for reinforcement (e.g., VR 360).

78 Chapter 4

Some VR schedules exist naturally; others may be created deliberately. Con-
sider again the example of Paul, the man with intellectual disability who packages
parts in a factory.

Describe how a VR 20 schedule of reinforcement would be implemented with
Paul.

The supervisor could reinforce his work performance on a VR 20 schedule by
delivering a token after an average of 20 parts that Paul packages. Sometimes the
number of responses needed would be less than 20 and sometimes more than 20.
The number of responses needed for any particular token delivery would not be
predictable to Paul, in contrast with the FR 20 schedule, where the token is pro-
vided after every 20 responses (packaged parts). Another common example of a
VR schedule is the slot machine found in casinos. The response of putting a
coin in the machine and pulling the handle is reinforced on a VR schedule. The
gambler never knows how many responses are needed for a jackpot (the rein-
forcer). However, the more responses the gambler makes, the more likely a jack-
pot is (because a VR schedule is based on number of responses, not on time or
some other factor). Therefore, the VR schedule in a slot machine produces high,
steady rates of responding. Of course, the casino makes sure that the VR schedule
is such that gamblers put more money in the machine than the machine pays out
as reinforcers. One other example of a VR schedule can be found in the salesper-
son who must make calls (in person or on the phone) to sell products. The num-
ber of calls that must occur before a sale (the reinforcer) occurs is variable. The
more calls the salesperson makes, the more likely it is that a sale will result. How-
ever, which call will result in a sale is unpredictable.

In the FR and VR schedules, the delivery of the reinforcer is based on the num-
ber of responses that occur. As a result, in both FR and VR schedules, more frequent
responding results in more frequent reinforcement. That is why ratio schedules are
the intermittent schedules most often used in behavior modification procedures.

Fixed Interval
With interval schedules (fixed interval, variable interval), a response is reinforced
only if it occurs after an interval of time has passed. It does not matter how many
responses occur; as soon as the specified interval of time has elapsed, the first
response that occurs is reinforced. In a fixed interval (FI) schedule, the interval
of time is fixed, or stays the same each time. For example, in a fixed interval
20-second (FI 20-second) schedule of reinforcement, the first response that occurs
after 20 seconds has elapsed results in the reinforcer. Responses that occur before
the 20 seconds are not reinforced; they have no effect on the subsequent delivery
of the reinforcer (i.e., they don’t make it come any sooner). Once the 20 seconds
has elapsed, the reinforcer is available, and the first response that occurs is rein-
forced. Then, 20 seconds later, the reinforcer is available again, and the first
response that occurs produces the reinforcer. Consider again the example of
Paul, who packages parts in a factory.

Describe how an FI 30-minute schedule of reinforcement would be implemented
with Paul.

Reinforcement 79

An FI 30-minute schedule would be in effect if the supervisor came by once
every 30 minutes and gave Paul a token for the first response (packaging a part)
that occurred. The number of parts that Paul packaged throughout the 30 min-
utes would be irrelevant. The supervisor would provide the token (reinforcer) for
the first part that she saw Paul package after the 30-minute interval. This is differ-
ent from an FR or VR schedule, in which Paul gets a token for the number of
parts he packages. In an FI schedule, only one response is needed for reinforce-
ment, but it must occur after the interval.

What Ferster and Skinner (1957) found is that FI schedules of reinforcement
produced a certain pattern of responding: The pigeon made an increasing num-
ber of responses near the end of the interval, up until the reinforcer was delivered.
After that, there was a pause in responding; as the end of the interval approached,
the pigeon again started responding more quickly until the reinforcer was deliv-
ered. We might expect to see the same pattern of behavior with Paul in the fac-
tory. After he receives the token from the supervisor and the supervisor walks
away (to observe other workers), Paul may slow down or stop working for a
while, and then start working again as the end of the 30 minutes approaches.
Because he receives a token for packaging a part only after the 30-minute interval
has ended, his behavior of packaging parts naturally starts to occur more fre-
quently as the end of the interval approaches. Because he never receives a token
for packaging parts during the 30-minute interval, his behavior naturally starts to
occur less frequently in the early part of the interval. This pattern of behavior (a
greater rate of responding near the end of the interval) is characteristic of FI sche-
dules of reinforcement. For this reason, FI schedules rarely are used in teaching
or training programs. Instead, FR or VR schedules are more commonly used
because they produce higher and steadier rates of responding. With an FR or VR
schedule, Paul learned to package more parts to receive more tokens. With an FI
schedule, Paul learned to package parts in a limited period around the end of
each 30-minute interval.

Variable Interval
In a variable interval (VI) schedule of reinforcement, as in an FI schedule, the
reinforcer is delivered for the first response that occurs after an interval of time
has elapsed. The difference is that in a VI schedule, each time interval is a differ-
ent length. The interval varies around an average time. For example, in a variable
interval 20-second (VI 20-second) schedule, sometimes the interval is more than
20 seconds and other times it is less than 20 seconds. The interval length is not
predictable each time, but the average length is 20 seconds. Ferster and Skinner
(1957) investigated various VI schedules of reinforcement. They found that the
pattern of responding on a VI schedule was different from that on an FI schedule.
On the VI schedule, the pigeon’s behavior (pecking the key) occurred at a steady
rate, whereas on the FI schedule, the frequency decreased in the early part of the
interval and increased near the end of the interval. Because the length of the
interval—and thus the availability of the reinforcer—was unpredictable in a VI
schedule, this off-and-on pattern of responding did not develop.

Once again, consider the case of Paul packaging parts in a factory.

80 Chapter 4

Describe how the supervisor would implement a VI 30-minute schedule with
Paul. Describe how Paul’s behavior would be different on a VI 30-minute sched-
ule from his behavior on an FI 30-minute schedule.

Using a VI 30-minute schedule, the supervisor would come around at
unpredictable intervals of time (e.g., after 5 minutes, 22 minutes, 45 minutes,
36 minutes) and give Paul a token for the first part that she saw Paul package.
The various intervals of time would average 30 minutes. The reinforcer (token)
would be given for the first response after the interval. On a VI 30-minute
schedule, Paul probably would package parts more steadily throughout the day.
The slowing down and speeding up of his work rate observed on the FI
30-minute schedule would not occur because the length of the intervals is
unpredictable.

Schedules of Reinforcement

Fixed ratio Reinforcer delivered after a certain number of responses. Produces high rate of
behavior, with a pause after reinforcement.

Variable ratio Reinforcer delivered after an average of x responses. Produces a high and steady
rate of behavior, with no pause after reinforcement.

Fixed interval Reinforcer delivered for the first response that occurs after a fixed interval of time.
Produces a low rate of behavior, with an on-and-off pattern. The response rate
increases near the end of the interval.

Variable interval Reinforcer delivered for the first response that occurs after a variable interval of
time. Produces a steady, low-to-moderate rate of behavior, with no on-and-off
pattern.

Reinforcing Different Dimensions
of Behavior

Although reinforcement often is used to increase the rate of a behavior, reinforce-
ment may also influence other dimensions of a behavior such as duration, inten-
sity, or latency. If a reinforcer is contingent on a particular duration of a
behavior, that duration of the behavior is more likely to occur. For example, if
a child is allowed to go outside and play after school only after she completes a
half hour of homework, she will be more likely to work on her homework for
30 minutes. Likewise, if the reinforcer is contingent on a particular intensity of a
behavior, the behavior is more likely to occur with that intensity. For example, if
a door gets stuck in cold weather and you must push harder to open it, then push-
ing harder is reinforced and you are more likely to push harder (increased inten-
sity) to open the door. Likewise, if a reinforcer is contingent on decreasing the
latency of a response, then decreased latency (increased speed) is strengthened.
For example, if a child receives a reinforcer for complying with a parent’s instruc-
tion immediately after the instruction is given, then an immediate response (short
latency) is strengthened and the child is more likely to respond immediately when
the parent makes a request.

Reinforcement 81

Concurrent Schedules of Reinforcement

In most situations, it is possible for a person to engage in more than one behavior.
For each of the possible behaviors a person could engage in at a particular time,
there is a specific schedule of reinforcement. All of the schedules of reinforce-
ment that are in effect for a person’s behaviors at one time are called concurrent
schedules of reinforcement. In other words, a number of different behaviors or
response options are concurrently available for the person. These are called
concurrent operants. Concurrent schedules of reinforcement (and punishment)
for the different response options at a particular time influence the probability
that a particular behavior will occur at that time. The person typically will engage
in one of the response options depending on the schedule of reinforcement, the
magnitude of reinforcement, the immediacy of reinforcement, and the response
effort for the various response options (Neef, Mace, & Shade, 1993; Neef, Mace,
Shea, & Shade, 1992; Neef, Shade, & Miller, 1994). For example, if Rayford had
the opportunity to do yard work for his friend for $10.00 per hour or to help his
cousin at the hardware store for $8.00 an hour, he probably would help his friend
because the magnitude of reinforcement is greater. If both jobs paid $10.00 per
hour but one job was much easier, Rayford probably would choose the easier
job. However, if he had the opportunity to spend the afternoon water-skiing with
his girlfriend, he might choose that over either job because it involved a more
powerful reinforcer than the amount of money earned from either job.

Research into concurrent schedules of reinforcement has shown that people
most often engage in the behavior that results in more frequent reinforcement, a
greater magnitude of reinforcement, more immediate reinforcement, or less response
effort (Friman & Poling, 1995; Hoch, McComas, Johnson, Faranda, & Guenther,
2002; Hoch, McComas, Thompson, & Paone, 2002; Neef et al., 1992, 1993, 1994;
Piazza, Roane, Keeney, Boney, & Abt, 2002). Information about concurrent sche-
dules is important in applying behavior modification because a schedule of reinforce-
ment for an undesirable behavior may exist concurrently with a schedule of
reinforcement for a desirable behavior. When using reinforcement to increase the
desirable behavior, you must also consider (and in some cases modify) the schedule
of reinforcement for the undesirable behavior (Mace & Roberts, 1993).

CHAPTER SUMMARY

1. Reinforcement is a basic principle of behavior.
Reinforcement is defined to occur when the
occurrence of a behavior is followed by an imme-
diate consequence that results in a strengthening
of the behavior or an increase in the probability of
the behavior in the future. Reinforcement is the
process responsible for the occurrence of operant
behavior.

2. Positive and negative reinforcement both strengthen
behavior. They differ only in whether the conse-
quence of the behavior is the addition of a stimulus

(positive reinforcer) or the removal of a stimulus
(aversive stimulus).

3. Unconditioned reinforcers are stimuli that are
naturally reinforcing because they have survival
value or biological importance. Conditioned rein-
forcers are originally neutral stimuli that have
been established as reinforcers because they
were paired with unconditioned reinforcers or
other conditioned reinforcers.

4. A number of factors influence the effectiveness of
reinforcement. A reinforcer should be delivered

82 Chapter 4

immediately to be most effective. A reinforcer is
most effective when it is contingent on the behav-
ior, that is, when it is delivered only if the behav-
ior occurs. Reinforcers are effective when there is
a state of deprivation or some other establishing
operation in effect. Generally, a reinforcer is more
effective the larger its amount or magnitude.

5. Reinforcement may be scheduled to occur every
time the behavior occurs (continuous reinforce-
ment [CRF]), or it may occur intermittently.
CRF schedules are used for acquisition, that is,
learning a new behavior. Intermittent schedules
are used to maintain the occurrence of a behavior
once it has been learned. There are four basic
intermittent reinforcement schedules. In ratio
schedules, a number of responses must occur for

the reinforcer to be delivered. In an FR schedule,
the number of responses is fixed, or constant; in a
VR schedule, the number of responses required
for reinforcement varies around an average num-
ber. In interval schedules, an interval of time must
pass before a response is reinforced. In an FI
schedule, the interval is fixed; in a VI schedule,
the interval varies around an average time. Ratio
schedules produce the highest rate of responding,
although there is often a pause after reinforce-
ment in FR schedules. Interval schedules produce
lower rates of responding than do ratio schedules.
The VI schedule produces a steady rate, whereas
the FI schedule produces an off-and-on pattern of
responding in which most responses occur as the
end of the interval approaches.

KEY TERMS

abolishing operation, 73
acquisition, 77
aversive stimulus, 66
avoidance behavior, 69
backup reinforcer, 71
concurrent operants, 82
concurrent schedules of

reinforcement, 82
conditioned reinforcer, 71
consequence, 61
contingency, 72
continuous reinforcement

schedule (CRF) schedule, 76
deprivation, 73

escape behavior, 69
establishing operation, 73
fixed interval schedule, 79
fixed ratio (FR) schedule, 78
generalized conditioned

reinforcer, 71
intermittent reinforcement

schedule, 76
maintenance, 77
motivating operations, 73
negative reinforcement, 66
operant behavior, 63
positive reinforcement, 66
positive reinforcer, 66

Premack principle, 68
Reinforcement, 61
reinforcer, 63
response, 72
response effort, 82
satiation, 74
schedule of reinforcement, 76
stimulus, 66
token, 71
unconditioned reinforcer, 71
variable interval (VI) schedule, 80
variable ratio (VR) schedule, 78

PRACTICE TEST

1. What is the definition of reinforcement?
(pp. 61, 63)

2. What was the reinforcer for Thorndike’s cat? What
behavior resulted in the reinforcer? What effect did
reinforcement have on the cat’s behavior? (p. 61)

3. What does it mean to say that a behavior is
strengthened? (p. 61)

4. What is an operant behavior? What operant
behavior of the rat was reinforced in Skinner’s
experiments? (p. 63)

5. Draw a graph that shows the effect of reinforce-
ment on the duration of cooperative play in a
child (p. 64).

6. Provide a definition of positive reinforcement.
(p. 66)

7. Provide a definition of negative reinforcement.
(p. 66)

8. Provide a novel example (not from the chapter)
of positive reinforcement.

Reinforcement 83

9. Provide a novel example of negative reinforce-
ment.

10. In what way are positive and negative reinforce-
ment alike? How are they different? (p. 66)

11. How is negative reinforcement different from
punishment? (p. 67)

12. What is an aversive stimulus? Provide an exam-
ple. (p. 66–67)

13. What is an unconditioned reinforcer? Provide
examples of unconditioned reinforcers. (p. 71)

14. What is a conditioned reinforcer? Provide exam-
ples. How did the stimulus in each example
become a conditioned reinforcer? (p. 71)

15. Identify the five factors that influence the effec-
tiveness of reinforcement. (pp. 72–76)

16. What is meant by contiguity between a response
and a reinforcer? How does contiguity influence
the effectiveness of reinforcement? (p. 72)

17. What is a reinforcement contingency? How does
a contingency influence the effectiveness of
reinforcement? (p. 72–73)

18. What is an establishing operation? What is an
abolishing operation? Provide some examples.
(pp. 73–74)

19. How can you determine whether a particular
stimulus is a reinforcer for a particular person?
(p. 75)

20. Distinguish between intermittent and continu-
ous schedules of reinforcement. (p. 76)

21. A CRF schedule is used for acquisition, and an
intermittent schedule is used for maintenance.
Describe what this means. (p. 76)

22. What is a fixed ratio schedule? A variable ratio
schedule? Provide an example that illustrates
each schedule. (pp. 78–79)

23. What is a fixed interval schedule? A variable
interval schedule? Describe the pattern of
responding you would expect with a fixed inter-
val schedule. (pp. 79–81)

24. Are interval or ratio schedules more likely to
be used in teaching or training programs?
Why? (p. 79)

25. What are concurrent schedules of reinforce-
ment? Provide an example. (pp. 82)

26. Identify each of the following as an example of
positive reinforcement or negative reinforcement.
a. Althea interrupts her parents and they scold

her each time she interrupts. Althea con-
tinues to interrupt her parents.

b. Dick curses at his teacher whenever she asks
him to do his math problems. The teacher
sends him to sit at the back of the room by
himself for 15 minutes (and thus he gets out
of doing his math problems) when he curses
at her. He continues to curse at her when she
tells him to do his math problems.

c. Maxine has a bad rash. Whenever she scratches
it, the itching goes away for a while. Maxine
continues to scratch her rash when it itches.

d. Jorge handed in his homework on time and
his teacher smiled at him. As a result, he
continues to hand in his homework on time.

e. Wiley drives his pickup truck fast down a dirt
road and spins around in the mud. As a
result, he is more likely to drive his pickup
truck fast down the dirt road.

f. Marcia’s mother yells at her when she doesn’t
clean her roomon Saturday.Asaresult,Marcia is
more likely to stay at a friend’s house on Saturday
so as to avoid being yelled at by her mother.

APPENDIX A
Operant Behaviors and Reinforcers from Each Example in Table 4-1

Operant Behavior Reinforcer

1. Child’s crying Parents’ attention

2. Opening the umbrella Keeps rain from falling on her

3. Turning on the fan Removes smoke from the kitchen

4. Asking her friend for the answer to a study question Friend provides correct answer

84 Chapter 4

5. Johnny sits in his seat Teacher smiles and praises him

6. Patricia puts in earplugs TV noise is eliminated

7. Employees assemble bicycles They earn money

8. Child’s tantrum Receipt of candy

9. Mother gives child candy Tantrum stops

APPENDIX B
Examples of Positive Reinforcement and Negative Reinforcement from Table 4-1

1. Positive reinforcement. The parents’ attention is a
positive reinforcer for the child’s crying. (The ces-
sation of crying is also negatively reinforcing for
the parents’ behavior of providing attention to
their child when she cries.)

2. Negative reinforcement. Opening the umbrella
prevents the rain from hitting the woman’s head
(removes an aversive stimulus).

3. Negative reinforcement. Turning on the exhaust
fan removes the smoke.

4. Positive reinforcement. Her friend provides the
correct answer to the question when the student
asks her for the correct answer.

5. Positive reinforcement. The teacher’s smile and
praise are a positive reinforcer for Johnny’s sitting
and paying attention.

6. Negative reinforcement. Putting in earplugs results
in the termination of the TV noise.

7. Positive reinforcement. Money is a positive rein-
forcer for assembling bicycles.

8. Positive reinforcement for the child’s behavior.
Getting candy from his mother reinforces the
child’s tantrum behavior. Negative reinforcement
for the mother’s behavior. The termination of
the child’s tantrum reinforces the mother’s behav-
ior of giving the child candy.

Reinforcement 85

This page intentionally left blank

Chapter Five

Extinction

As discussed in Chapter 4, reinforcement is responsible for the acquisition andmaintenance of operant behavior. This chapter discusses extinction, a pro-
cess that weakens operant behavior. Consider the two examples that follow.

Every Monday, Wednesday, and Friday, Rae walks to her behav-
ior modification class at 8 A.M. Right before class each day, she stops
at the coffee machine, puts a dollar in the machine, pushes the but-
ton, and gets her coffee for class. One day, she walks up to the
machine, puts her money in, and pushes the button, but nothing hap-
pens. She pushes the button again, and nothing happens. She pushes
the button harder and harder and then slams the button a few times,
but she still does not get her coffee. Finally, she gives up and walks to
class without her coffee. She doesn’t try the machine again for a
week, but then she tries it again, and the same thing happens. From
then on, she never tries the machine again; instead, she gets her cof-
fee at the convenience store on the way to school.

Each evening when Greg gets home from work, he goes into his apartment build-
ing through the emergency exit in back because that door is close to his apartment and
he doesn’t have to walk all the way around to the front door. The apartment manager
doesn’t want people to use this door except in emergencies, so she installs a new lock
on the door. That day, when Greg gets home from work, he turns the doorknob but
the door doesn’t open. He turns the knob again, but nothing happens. He starts turning
the knob harder and pulling harder on the door, but still nothing happens. Eventually
he stops and walks to the front door. Greg tries the door again the next couple of days
when he gets home from work, but still it will not open. Finally, he quits trying to go in
through the emergency door.

Response

Rae puts money in coffee machine. No coffee comes out of coffee machine.

Outcome: Rae is less likely to put money in the coffee machine in the future.

â–  What is the principle of extinction?

â–  What happens during an extinction
burst?

â–  How is extinction different after
positive reinforcement and negative
reinforcement?

â–  What is a common misconception
about extinction?

â–  What factors influence extinction?

87

Defining Extinction

The basic behavioral principle that is illustrated in the preceding examples is
extinction. In each example, a behavior that had been reinforced for a period of
time was no longer reinforced and, therefore, the behavior stopped occurring. Rae’s
behavior of putting money in the coffee machine and pushing the button was rein-
forced by getting coffee. Greg’s behavior of turning the doorknob and opening the
emergency door was reinforced by entering his apartment building at a point closer
to his apartment. These behaviors were reinforced on a continuous reinforcement
schedule. Once the reinforcement stopped, Rae and Greg both engaged in the
behavior less and less, and ultimately stopped engaging in the behavior.

Extinction is a basic principle of behavior. The behavioral definition of
extinction is as follows: Extinction occurs when

1. A behavior that has been previously reinforced
2. no longer results in the reinforcing consequences
3. and, therefore, the behavior stops occurring in the future.

As long as a behavior is reinforced, at least intermittently, it will continue to
occur. If a behavior is no longer followed by a reinforcing consequence, however,
the person will stop engaging in the behavior. When a behavior stops occurring
because it is no longer reinforced, we say that the behavior has undergone extinc-
tion or that the behavior has been extinguished.

Skinner (1938) and Ferster and Skinner (1957) demonstrated the principle of
extinction with laboratory animals. When the pigeon in the experimental chamber
no longer received food as a reinforcer for pecking the key, the pigeon’s key-pecking
behavior stopped. When the laboratory rat no longer received food pellets for press-
ing the lever, the lever-pressing behavior decreased and eventually stopped.

Of course, numerous research studies have also demonstrated the principle of
extinction with human behavior (see Lerman & Iwata, 1996b). In one of the earliest
studies reporting the use of extinction to decrease a problem behavior, Williams (1959)
illustrated the effectiveness of extinction in decreasing the nighttime tantrums of a young
child. Because Williams had determined that the child’s tantrum behavior was being
reinforced by the parents’ attention, the extinction procedure called for the parents to
refrain from providing attention when the child engaged in tantrum behaviors at night.

Response

Child tantrums at bedtime. Parent does not pay attention to the child.

Outcome: In the future, child is less likely to engage in tantrum behavior at bedtime.

Response

Greg turns the handle on the emergency exit door. The door does not open.

Outcome: In the future, Greg is less likely to try to open the emergency exit door.

88 Chapter 5

Numerous studies have demonstrated the effectiveness of extinction for decreas-
ing problem behaviors in children and adults (Ayllon & Michael, 1959; Ducharme &
Van Houten, 1994; Holz, Azrin, & Ayllon 1963; Lerman & Iwata, 1995; Mazaleski,
Iwata, Vollmer, Zarcone, & Smith, 1993; Neisworth & Moore, 1972; Rincover, 1978;
Wright, Brown, & Andrews, 1978). In each of these studies, the reinforcer for a
problem was eliminated or withheld following the occurrence of the behavior, and the
behavior decreased in the future. Consider the study by Hasazi and Hasazi (1972),
who used extinction to reduce arithmetic errors made by an 8-year-old boy. Whenever
the boy did addition problems with two-digit answers, he reversed the digits (e.g.,
he wrote 21 instead of 12 as the answer to 7 þ 5). The researchers determined that the
attention (extra help) provided by the teacher for incorrect answers was reinforcing
the child’s behavior of reversing the digits. The extinction procedure required
the teacher to refrain from providing attention for incorrect answers. The teacher
also praised the child for correct answers. (This is differential reinforcement; see
Chapter 15 for a more detailed discussion.) The child’s digit-reversal behavior
decreased dramatically when extinction was implemented (Figure 5-1). This study
is particularly interesting because many professionals would have considered the digit
reversal to be a sign of a learning disability, whereas the authors demonstrated that the
digit reversal actually was an operant behavior reinforced by the teacher’s attention.

In another example, Lovaas and Simmons (1969) used extinction to reduce the
self-injurious behavior of a child with intellectual disability. Lovaas and Simmons
believed the child’s head-hitting behavior was being reinforced by social consequences
(attention) from adults. Extinction therefore involved removing adult attention when-
ever the child hit himself. The results showed that the frequency of head-hitting
decreased from more than 2500 hits in a 1-hour session to zero per session. It took
ten sessions of extinction for the frequency of the behavior to decrease to zero.

N
um

be
r o

f D
ig

it
Re

ve
rs

al
s

4

8

12

16

20
Baseline 1 Extinction 1 Baseline 2 Extinction 2

Days
4 8 12 16 20 24 28

••
•• ••

•
•

•

•
• •

•

•
•

•• ••••

•• •• ••

FIGURE 5-1 This graph shows the effect of extinction on the digit-reversal behavior of an 8-year-old boy.
The graph illustrates an A-B-A-B reversal design. During baseline, the behavior of making digit
reversals in his answers to addition problems is reinforced by the teacher’s attention. When making
digit reversals was no longer reinforced by teacher attention, the frequency of the behavior
decreased dramatically. (From Hasazi, J. E., & Hasazi, S. E. [1972]. Effects of teacher attention on
digit reversal behavior in an elementary school child. Journal of Applied Behavior Analysis, 5,
157–162. Copyright © 1972 University of Kansas Press. Reprinted by permission of Joseph E.
Hasazi and the publisher.)

Extinction 89

On Terms: Talking About Extinction Correctly

When talking about extinction or the effects of extinction:

■ It is correct to say you extinguished a behavior; it is NOT correct to say you “extincted” a behavior

â–  It is correct to say you put a behavior on extinction; it is NOT correct to say you put a person on
extinction

â–  It is correct to say a behavior is extinguished; it is NOT correct to say a behavior is extinct.

Extinction Burst

One characteristic of the extinction process is that once the behavior is no longer
reinforced, it often increases briefly in frequency, duration, or intensity before it
decreases and ultimately stops (Lerman & Iwata, 1995). In the first example,
when Rae did not get her coffee, she pushed the button on the coffee machine
repeatedly (increase in frequency), and then pushed it harder and harder (increase
in intensity) before finally giving up. When Greg found the back door to his apart-
ment building would not open, he turned the handle and pulled the doorknob a
number of times (increase in frequency), and he pulled harder on the doorknob
(increase in intensity) before finally giving up. Increase in frequency, duration, or
intensity of the unreinforced behavior during the extinction process is called an
extinction burst. Consider two other examples.

When Mark pushes the on button on the remote control for his TV set and it
does not turn on the TV (because the batteries are dead), he pushes it longer
(increased duration) and harder (increased intensity) before he finally gives up.
His behavior of pushing the on button was not reinforced by the TV turning on;
therefore, he quit trying, but not until he tried pushing it longer and harder
(extinction burst).

Each night, 4-year-old Amanda cried at bedtime for 10-15 minutes, and her
parents came to her room and talked to her until she fell asleep. By doing so, her
parents were accidentally reinforcing her crying. After talking to her pediatrician,
the parents decided not to go into her room or talk to her when she cried at bed-
time. The first night, she cried for 25 minutes before falling asleep. By the end of
the week, she quit crying at all at bedtime. When they stopped going to her room
after she cried, the parents were using extinction. The increase in crying duration
the first night is an extinction burst. Figure 5-2 shows the graph of Amanda’s cry-
ing duration before and after her parents used extinction. Once the parents imple-
mented extinction, the behavior increased briefly, but then decreased and
eventually stopped altogether.

Response

Outcome: The child is less likely to hit himself in the head because the behavior is no longer reinforced by
adult attention.

The child hits his head. He receives no attention from adults.

90 Chapter 5

One other characteristic of an extinction burst is that novel behaviors (behaviors
that do not typically occur in a particular situation) may occur for a brief period when
a behavior is no longer reinforced (Bijou, 1958; Lalli, Zanolli, & Wohn, 1994). For
example, when Amanda’s parents no longer reinforced her crying at night, she cried
longer and louder (increased duration and intensity), but she also screamed and hit
her pillow (novel behaviors). In the first example, Rae not only pushed the button
on the coffee machine repeatedly when the coffee didn’t come out, but she also
pushed the coin return button and shook the machine (novel behaviors; Figure 5-3).

Days

Du
ra

tio
n

of
C

ry
in

g
(m

in
)

0

5

10

15

20

25

30

0 102 4 6 8 1412

Baseline Extinction

• • •

• • •

• •

•

• • • •

•

FIGURE 5-2 The graph shows hypothetical data on the duration of crying during baseline and extinction. On the
first day of extinction, an extinction burst occurred: The behavior increased in duration. On subse-
quent days, it decreased and eventually stopped.

FIGURE 5-3 When Rae does not get her coffee from the vending machine, she pushes the buttons repeatedly
and shakes the machine. This is an example of an extinction burst.

Extinction 91

Sometimes, the novel behaviors during extinction bursts may include emo-
tional responses (Chance, 1988). For example, Rae may act in an angry fashion
and curse at the coffee machine or kick it. Azrin, Hutchinson, and Hake (1966)
reported that aggressive behavior often is seen when extinction is used. It is not
uncommon for young children to exhibit emotional responses when their behav-
ior is no longer reinforced. The child whose request for candy is denied may
scream and cry. The parent may then unwittingly reinforce this screaming and
crying by giving the child some candy. As you may recall from Chapter 4, the par-
ent’s behavior of giving candy to the child is negatively reinforced by the termina-
tion of the child’s screaming and crying.

The extinction burst, which involves an increase in the unreinforced behavior
or the occurrence of novel (and sometimes emotional) behaviors for a brief
period, is a natural reaction to the termination of reinforcement. The increased
frequency, duration, or intensity of the unreinforced behavior—or the novel beha-
viors that occur during extinction—may be reinforced; thus, the extinction burst
serves a valuable purpose. For example, when Greg tugs very hard on the door-
knob, it may open for him if it is only stuck, rather than locked. When Amanda
screams and cries louder, her parents may come into the room and give her the
attention she wasn’t getting for simply crying.

The extinction burst is not necessarily a conscious process, however. Amanda
probably is not thinking, “I’ll cry louder, scream, and hit my pillow to get my par-
ents’ attention.” The extinction burst is simply a natural characteristic of an extinc-
tion situation.

Extinction Burst

When a behavior is no longer reinforced, three things may happen.

â–  The behavior may briefly increase in frequency, duration, or intensity.

â–  Novel behaviors may occur.

â–  Emotional responses or aggressive behavior may occur.

FOR FURTHER READING
Extinction Burst

The extinction burst is an important phenomenon with direct implications for the use of extinction to
decrease a behavior problem (see Chapter 14). The extinction burst has been studied by a number of
researchers. For example, Lerman and Iwata (1995) reviewed published studies evaluating extinction
and found that an extinction burst was evident in 24% of the studies. They identified an extinction
burst as an initial increase in the behavior during extinction. Lerman, Iwata, and Wallace (1999) exam-
ined the use of extinction in 41 cases of self-injurious behavior over 9 years in their own treatment pro-
gram. They found evidence of an extinction burst (initial increase in the behavior) in 39% of the cases
and an increase in aggressive behavior in 22% of the cases. Interestingly, an extinction burst was more
likely to occur after extinction of a negatively reinforced behavior than extinction of a positively reinforced
behavior. In both studies, an extinction burst was more likely when extinction was used alone than when
extinction was used with other treatments.

92 Chapter 5

Spontaneous Recovery

One other characteristic of extinction is that the behavior may occur again even
after it has not occurred for some time. This is called spontaneous recovery.
Spontaneous recovery is the natural tendency for the behavior to occur again in
situations that are similar to those in which it occurred and was reinforced before
extinction (Chance, 1988; Lerman, Kelly, Van Camp, & Roane, 1999; Zeiler,
1971). If extinction is still in place when spontaneous recovery occurs—that is, if
there is no reinforcement—the behavior will not continue for long. Once in a
while, Amanda may cry at night long after extinction, but if she gets no attention
for the crying, it will not occur often or for very long. However, if spontaneous
recovery occurs and the behavior is now reinforced, the effect of extinction will
be lost. For example, Greg may still try occasionally to open the back door to his
apartment building. If the door happens to open one day, his behavior of using
that door will be reinforced, and he will be more likely to try to use that door
again. Finding the door open occasionally would be an example of intermittent
reinforcement, which would increase behavioral persistence or resistance to
extinction in the future.

Procedural Variations of Extinction

As discussed in Chapter 4, there are two procedural variations or types of rein-
forcement: positive reinforcement and negative reinforcement.

Define positive reinforcement. Define negative reinforcement.

A behavior may undergo extinction regardless of whether it is maintained by
positive or negative reinforcement. The outcome of extinction is the same: The
behavior decreases and stops occurring. Procedurally, however, extinction is
slightly different in the two cases. If a behavior is positively reinforced, a conse-
quence is applied or added after the behavior. Therefore, extinction of a positively
reinforced behavior involves withholding the consequence that was previously
delivered after the behavior. To put it another way, when the behavior no longer
results in the delivery of the reinforcing consequence, the behavior no longer
occurs.

If a behavior is negatively reinforced, the behavior results in the removal or
avoidance of an aversive stimulus. Extinction of a negatively reinforced behavior
therefore involves eliminating the escape or avoidance that was reinforcing the
behavior. When the behavior no longer results in escape from or avoidance of an
aversive stimulus, the behavior eventually stops. For example, suppose that you
wear earplugs in your job at the factory to decrease the loud noise of the equip-
ment. Wearing earplugs is negatively reinforced by escape from the loud noise. If
the earplugs wore out and they no longer decreased the noise, you would stop
wearing them. The behavior of wearing earplugs would be extinguished because
wearing them no longer produced escape from the noise. This may be a difficult
concept to grasp. Consider the following examples.

Extinction 93

Procedural Variations of Extinction
â–  The positive reinforcer is no longer delivered after the behavior

â–  The aversive stimulus is no longer removed after the behavior

Shandra has an 11 P.M. curfew. If she comes in later than 11 P.M., her par-
ents scold her, lecture her, and ground her for a week. Because the parents go
to bed at 10 P.M., they do not know what time their daughter comes home.
They ask her the next morning, and if she came home after 11 P.M., she lies
and tells them she was home earlier. Lying is negatively reinforced by the avoid-
ance of aversive consequences from her parents. Extinction of lying would occur
if lying no longer helped her to avoid aversive consequences. Thus, if a parent
were awake in bed and knew when Shandra came home, she would not avoid
aversive consequences by lying. As a result, she would quit lying when she got
home late.

Consider another example. Joe is a college student who works part-time as a
custodian. He hates to clean the bathrooms. Whenever the supervisor asks Joe to
do so, Joe makes up excuses why he can’t clean the bathrooms, and the supervisor
lets him out of the job and asks someone else to do it. Joe’s behavior of making up
excuses helps him avoid cleaning bathrooms. Making up excuses, therefore, is
negatively reinforced.

How would the supervisor use extinction to stop Joe from making excuses?

Every time Joe makes up excuses, the supervisor tells him to clean the bath-
room anyway. Therefore, when Joe cannot avoid cleaning bathrooms by making
up excuses, he will stop making up excuses.

Response

Outcome: Shandra is more likely to lie about coming home late in the future.

Shandra lies to her parents when she comes home after curfew. Shandra avoids getting scolded and grounded.

ReinforcementReinforcement

Response

Outcome: Shandra is less likely to lie about coming home late in the future.

Shandra lies to her parents when she comes home after curfew. The parents scold her and ground her.
She does not avoid aversive consequences.

ExtinctionExtinction

94 Chapter 5

Research by Brian Iwata and his colleagues (Iwata, Pace, Cowdery, &
Miltenberger, 1994) has demonstrated that extinction is procedurally different
when a behavior has been maintained by positive reinforcement than by negative
reinforcement. Iwata and his colleagues studied self-injurious behavior (such as
self-hitting) exhibited by children who were intellectually disabled. When they
found that self-injury was positively reinforced by attention from adults, they
implemented extinction by removing the adult attention after the behavior. For
some children, however, self-injury was negatively reinforced: The self-injurious
behavior resulted in escape from academic tasks. That is, a teacher stopped mak-
ing demands on a child (removed the academic demand) once the child started
to engage in self-injurious behavior. In these cases of negative reinforcement,
extinction required the teacher not to remove the academic demand after the
self-injury. Therefore, the self-injurious behavior no longer resulted in escape
from the teaching situation. Iwata and his colleagues clearly demonstrated that if
extinction is to occur, the reinforcer for the behavior must be identified, and that
particular reinforcer must be eliminated. Unless the appropriate reinforcer is iden-
tified and eliminated, the process does not function as extinction.

Edward Carr and his colleagues (Carr, Newsom, & Binkoff, 1980) studied the
behavior disorders of children with intellectual disabilities. They showed that
aggressive behavior in two children occurred only in demand situations and func-
tioned as escape behavior. That is, the aggressive behavior was negatively rein-
forced by the termination of demands.

How would extinction be used with the aggressive behavior of these two children?

Carr and his colleagues demonstrated that when the child could not escape
from the demand situation by engaging in aggressive behavior, the aggressive behav-
ior decreased dramatically. Because escape was reinforcing the aggressive behavior,
preventing escape functioned as extinction.

Response

Outcome: Joe is more likely to make excuses when asked to clean the bathroom in the future.

Joe makes excuses when asked to clean the bathroom. Joe avoids cleaning the bathroom.

ReinforcementReinforcement

Response

Outcome: Joe is less likely to make excuses when asked to clean the bathroom in the future.

Joe makes excuses when asked to clean the bathroom. The supervisor does not let him avoid cleaning the bathroom.

ExtinctionExtinction

Extinction 95

A Common Misconception about Extinction

Although extinction is procedurally different depending on the type of reinforce-
ment for the behavior, the outcome is always the same: The behavior stops. A com-
mon misconception is that using extinction simply means ignoring the behavior.
This is inaccurate in most cases. Extinction means removing the reinforcer for a
behavior. Ignoring the problem behavior functions as extinction only if attention is
the reinforcer. For example, a person’s shoplifting is reinforced by getting merchan-
dise from a store. If the salespeople in the store ignore the shoplifting behavior, this
will not cause that behavior to stop. Again, suppose that a child runs from the table
whenever he is told to eat his vegetables, and the outcome is that he does not eat his
vegetables. If the parents ignore this behavior, it will not stop. Running from the
table is reinforced by escape from eating the vegetables. Ignoring the behavior does
not take away this reinforcer and, therefore, does not function as extinction.

Take each example of reinforcement in Table 4-1 and turn it into an example of
extinction. Answers are provided at the end of this chapter in Appendix A.

Factors That Influence Extinction

Two important factors influence the extinction process: the reinforcement schedule
before extinction and the occurrence of reinforcement after extinction. The rein-
forcement schedule partly determines whether extinction results in a rapid decrease
in the behavior or a more gradual decrease (Bijou, 1958; Kazdin & Polster, 1973;
Lerman, Iwata, Shore, & Kahng, 1996; Neisworth, Hunt, Gallop, & Madle, 1985).
Recall from Chapter 4 that, in continuous reinforcement, every occurrence of a
behavior is followed by a reinforcer; in intermittent reinforcement, not every occur-
rence of a behavior results in a reinforcer; instead, the behavior is only occasionally
reinforced. When a behavior is continuously reinforced, it decreases rapidly once
the reinforcement is terminated. In contrast, when a behavior is intermittently rein-
forced, it often decreases more gradually once the reinforcement is terminated.
This occurs because the change from reinforcement to extinction is more discrimi-
nable (there is a larger contrast) when a behavior is reinforced every time than
when only some occurrences of the behavior result in reinforcement.

For example, if you put money into a vending machine and push the button,
you always get the item you want. This is a case of continuous reinforcement, and
the decrease in behavior during extinction would be fairly rapid. You would not
continue to put money into a vending machine if you no longer got the item
you paid for; the lack of reinforcement would be immediately apparent. Contrast
this with what happens when you put money into a slot machine or a video gam-
bling machine. This is a case of intermittent reinforcement: Putting money into
the slot machine is only occasionally reinforced by hitting the jackpot and win-
ning money from the machine. If the machine was broken and never again pro-
duced a jackpot (no reinforcement), you might put many more coins into the
machine before finally giving up. It takes longer for the gambling behavior to
stop because it is more difficult to determine that there is no longer reinforcement
for the behavior.

96 Chapter 5

Intermittent reinforcement before extinction produces resistance to extinction;
that is, the behavior persists once extinction is implemented. Continuous rein-
forcement before extinction produces much less resistance to extinction (less
behavioral persistence). Because of resistance to extinction, the reinforcement
schedule before extinction has implications for the successful use of extinction
in a behavior modification program (see Chapter 14).

A second factor that influences extinction is the occurrence of reinforcement
after extinction. If reinforcement occurs in the course of extinction, it takes longer
for the behavior to decrease. This is because reinforcement of the behavior, once
extinction has been started, amounts to intermittent reinforcement, which makes
the behavior more resistant to extinction. In addition, if the behavior is reinforced
during an episode of spontaneous recovery, the behavior may then increase to its
level before extinction. Consider the case of Amanda again. We see in Figure 5-2
that her crying at night decreased to zero by day 14, 7 days after extinction was
started. What if, on day 13, the babysitter came into her room and talked to her
when she cried that night? This would reinforce the crying behavior, and crying
would occur for many more days (Figure 5-4). The babysitter’s action (Figure 5-5)
would amount to intermittent reinforcement and would produce resistance to
extinction.

In the case of extinction of the child’s bedtime tantrum reported by Williams
(1959), the tantrums had almost stopped after the parents had used extinction for
a few days. However, when an aunt paid attention to the child’s tantrums one
night, they increased in intensity. Only when the parents again used extinction
consistently did the tantrums finally stop.

Days

D
ur

at
io

n
of

C
ry

in
g

(m
in

)

0

5

10

15

20

25

30

0 102 4 6 8 14 16 18 20 22 24 2612

Baseline Extinction

• • •

• • •

• •

•

•
•

•
• • • • •

•

• •

•

• • ••

FIGURE 5-4 The graph shows hypothetical data illustrating the duration of crying during baseline and extinction
if the behavior was accidentally reinforced on day 13. After day 13, the duration of the behavior
increased, and extinction was prolonged.

Extinction 97

CHAPTER SUMMARY

1. Extinction is a basic behavioral principle. It
occurs when a previously reinforced behavior is
no longer reinforced, and as a result, the behavior
decreases and stops occurring.

2. Extinction often is characterized by an extinction
burst, in which the unreinforced behavior tempo-
rarily increases in frequency, intensity, or dura-
tion, or novel behaviors or emotional responses
are exhibited temporarily.

3. Procedurally, extinction is different for behaviors
that are positively reinforced than for those that
are negatively reinforced. In each case, however,
the particular reinforcer for the behavior is termi-
nated, and the outcome is the elimination of the

behavior. With extinction of a positively reinforced
behavior, the positive reinforcer is no longer deliv-
ered after the behavior. With extinction of a nega-
tively reinforced behavior, the aversive stimulus is
no longer removed after the behavior.

4. A common misconception about extinction is that
extinction means ignoring the behavior. Ignoring
the behavior functions as extinction only if atten-
tion was the reinforcer for the behavior.

5. The behavior decreases more rapidly during
extinction if the behavior was reinforced on a con-
tinuous schedule before extinction and if the
behavior is never reinforced during the extinction
process.

KEY TERMS

extinction, 88
extinction burst, 90

resistance to extinction, 97
spontaneous recovery, 93

FIGURE 5-5 When Amanda cries at night, the babysitter comes into Amanda’s room and talks to her. By doing
so, the babysitter accidentally reinforces the crying. As a result, it will take longer for the behavior
to decrease and stop when the parents use extinction.

98 Chapter 5

PRACTICE TEST

1. What is the behavioral definition of extinction?
(p. 88)

2. Provide an example of extinction. (pp. 87–89)
3. What is an extinction burst? (p. 90)
4. Provide an example of an extinction burst.

(pp. 90–92)
5. Draw a graph of extinction. Be sure to show the

extinction burst. (p. 91)
6. What is negative reinforcement? Explain the

extinction of a negatively reinforced behavior.
(pp. 93–95)

7. Provide an example of extinction of a negatively
reinforced behavior. (pp. 93–95)

8. Extinction is not the same thing as ignoring.
Explain this statement. (p. 96)

9. Explain how the reinforcement schedule for a
behavior (continuous or intermittent) influences
extinction of the behavior. (pp. 96–97)

10. What happens to a behavior when it is acciden-
tally reinforced during the extinction process?
(p. 97)

11. Draw a graph of extinction that shows what hap-
pens when a behavior is accidentally reinforced.
(p. 97)

12. What is spontaneous recovery during extinction?
(p. 93)

APPENDIX A
Applying Extinction with Each Example of Reinforcement from Table 4-1

1. If the parents quit coming to the child’s room
when she cried at night, the child would stop cry-
ing in the future.

2. If the umbrella did not open correctly each time
the woman tried to open it, and as a result, it did
not keep the rain from hitting her, she would stop
using the umbrella in the future.

3. If the exhaust fan did not respond to the cook’s
attempts to turn it on, or if it did not effectively
draw the smoke out of the room, he would even-
tually stop trying to turn on the exhaust fan.

4. If the student’s roommate no longer gave her the
answers to the questions, she would stop asking
her roommate for the answers.

5. If the teacher ignored Johnny when he stayed in
his seat and paid attention in class, he would be

less likely to stay in his seat and pay attention in
the future.

6. If the earplugs no longer eliminated the noise
from the TV, Patricia would quit using them.

7. If the workers no longer earned money for assem-
bling bicycles (because the company was bank-
rupt), they would stop assembling bicycles for the
company.

8. If the mother did not give candy to her child
when he cried in the store, the child would be
less likely to cry in the store. If the child did not
stop crying when his mother gave him candy in
the store, the mother would be less likely to give
him candy when he cried, because giving him
candy was not reinforced by the termination of his
crying.

Extinction 99

This page intentionally left blank

Chapter Six

Punishment

In Chapters 4 and 5, we discussed the basic principles of reinforcement andextinction. Positive and negative reinforcement are processes that strengthen
operant behavior, and extinction is a process that weakens operant behavior. In
this chapter, we focus on punishment, another process that weakens operant
behavior (Lerman & Vorndran, 2002). Consider the following examples.

Kathy, a college senior, moved into a new apartment near campus. On her
way to class, she passed a fenced-in yard with a big, friendly looking dog. One

day, when the dog was near the fence, Kathy reached over to pet
the dog. At once, the dog growled, bared its teeth, and bit her
hand. After this, she never again tried to pet the dog.

On Mother’s Day, Otis decided to get up early and make break-
fast for his mom. He put the cast-iron skillet on the stove and turned
the burner on high. Then he mixed a couple of eggs in a bowl with
some milk to make scrambled eggs. After about 5 minutes, he
poured the eggs from the bowl into the skillet. Immediately, the
eggs started to burn and smoke rose from the skillet. Otis grabbed
the handle of the skillet to move it off of the burner. As soon as he
touched the handle, pain shot through his hand; he screamed and
dropped the skillet. After that episode, Otis never again grabbed the
handle of a hot cast-iron skillet. He always used a hot pad to avoid
burning himself.

Defining Punishment

The preceding two examples illustrate the behavioral principle of punishment.
In each example, a person engaged in a behavior and there was an immediate
consequence that made it less likely that the person would repeat the behavior in
similar situations in the future. Kathy reached over the fence to pet the dog, and
the dog immediately bit her. As a result, Kathy is less likely to reach over the
fence to pet that dog or other unfamiliar dogs. Otis grabbed the hot handle of a
cast-iron skillet, which resulted immediately in painful stimulation as he burned

â–  What is the principle of punishment?

â–  What is a common misconception
about the definition of punishment in
behavior modification?

â–  How does positive punishment differ
from negative punishment?

â–  How are unconditioned punishers
different from conditioned punishers?

â–  What factors influence the
effectiveness of punishment?

â–  What are the problems with
punishment?

101

his hand. As a result, Otis is much less likely to grab the handle of a cast-iron
skillet on a hot stove again (at least not without a hot pad).

As demonstrated in these examples, there are three parts to the definition of
punishment.

1. A particular behavior occurs.
2. A consequence immediately follows the behavior.
3. As a result, the behavior is less likely to occur again in the future. (The

behavior is weakened.)

A punisher (also called an aversive stimulus) is a consequence that makes a
particular behavior less likely to occur in the future. For Kathy, the dog bite was
a punisher for her behavior of reaching over the fence. For Otis, the painful stim-
ulus (burning his hand) was the punisher for grabbing the handle of the cast-iron
skillet. A punisher is defined by its effect on the behavior it follows. A stimulus or
event is a punisher when it decreases the frequency of the behavior it follows.

Consider the case of an aggressive and disruptive 5-year-old. Juan teases and
hits his sisters until they cry. His mother scolds him and spanks him each time
he teases or hits his sisters. Although Juan stops teasing and hitting his sisters at
the moment that his mother scolds him and spanks him, he continues to engage
in these aggressive and disruptive behaviors with his sisters day after day.

Is the scolding and spanking by his mother a punisher for Juan’s aggressive and
disruptive behavior? Why or why not?

No, the scolding and spanking do not function as punishers. They have not
resulted in a decrease in Juan’s problem behavior over time. This example actually
illustrates positive reinforcement. Juan’s behavior (teasing and hitting) results in the
presentation of a consequence (scolding and spanking by his mother and crying by
his sisters), and the outcome is that Juan continues to engage in the behavior day
after day. These are the three parts of the definition of positive reinforcement.

Response

Outcome: Kathy is less likely to reach over the fence in the future.

Kathy reaches over the fence and immediately the dog bites her.

Response

Outcome: Otis is less likely to grab a hot cast iron skillet in the future.

Otis touches a hot skillet and immediately he burns his hand (a painful stimulus).

102 Chapter 6

This raises an important point about the definition of punishment. You cannot
define punishment by whether the consequence appears unfavorable, unpleasant,
or aversive. You can conclude that a particular consequence is punishing only
if the behavior decreases in the future. In Juan’s case, scolding and spanking
appear to be unfavorable consequences, but he continues to hit and tease his
sisters. If the scolding and spanking functioned as a punisher, Juan would stop
hitting and teasinghis sisters over time.When wedefine punishment (or reinforcement)
according to whether the behavior decreases (or increases) in the future as a result of
the consequences, we are adopting a functional definition. See Table 6-1 for examples
of punishment.

TABLE 6-1 Examples for Self-Assessment (Punishment)

1. Ed was riding his bike down the street and looking down at the ground as he pedaled. All of a
sudden he ran into the back of a parked car, flew off the bike, and hit the roof of the car with his
face. In the process, he knocked his front teeth loose. In the future, Ed was much less likely to look
down at the ground when he rode his bike.

2. When Alma was in the day care program, she sometimes hit the other kids if they played with her
toys. Alma’s teacher made her quit playing and sit in a chair in another room for 2 minutes each
time she hit someone. As a result, Alma stopped hitting the other children.

3. Carlton made money in the summer by mowing his neighbor’s lawn each week. One week, Carlton ran
over the garden hose with the lawn mower and ruined the hose. His neighbor made Carlton pay for the
hose. Since then, whenever Carlton mows the lawn, he never runs over a hose or any other objects lying in
the grass.

4. Sarah was driving down the interstate on her way to see a friend who lived a few hours away. Feeling
a little bored, she picked up the newspaper on the seat next to her and began to read it. As she was
reading, her car gradually veered to the right without her noticing. Suddenly, the car was sliding
on gravel and sideswiped a speed limit sign. As a result, Sarah no longer reads when she drives on
the highway.

5. Helen goes to school in a special class for children with behavior disorders. Her teachers use poker
chips as conditioned reinforcers for her academic performance. The teachers place a poker chip in a
container to reinforce her correct answers. However, each time Helen gets out of her seat without
permission, the teachers take one chip away from her. As a result, Helen stopped getting out of her
seat without permission.

6. At parties, Kevin used to make jokes about his wife’s cooking and got a lot of laughs from his
friends. At first, his wife smiled at his jokes, but eventually she got upset; whenever Kevin made
a joke about her cooking, she gave him an icy stare. As a result, Kevin stopped joking about his
wife’s cooking.

Response

Outcome: Juan continues to hit and tease his sisters in the future.

Juan’s teasing and hitting scolding and spanking from his mother.is immediately followed by
his sisters’ crying

Punishment 103

One other point to consider is whether a behavior decreases or stops only
at the time the consequence is administered, or whether the behavior decreases
in the future. Juan stopped hitting his sisters at the time that he received a
spanking from his mother, but he did not stop hitting his sisters in the future.
Some parents continue to scold or spank their children because it puts an imme-
diate stop to the problem behavior, even though their scolding and spanking do
not make the child’s problem behavior less likely to occur in the future. The
parents believe they are using punishment. However, if the behavior continues
to occur in the future, the scolding and spanking do not function as punishers
and may actually function as reinforcers.

What reinforces the parents’ behavior of scolding and spanking the child?

Because the child temporarily stops the problem behavior after the scolding
or spanking, the parents’ behavior of scolding or spanking is negatively reinforced,
so the parents continue to scold or spank the child in the future when he or she
misbehaves.

A Common Misconception about
Punishment

In behavior modification, punishment is a technical term with a specific meaning.
Whenever behavior analysts speak of punishment, they are referring to a process
in which the consequence of a behavior results in a future decrease in the occur-
rence of that behavior. This is quite different from what most people think of as
punishment. In general usage, punishment can mean many different things, most
of them unpleasant.

Many people define punishment as something meted out to a person
who has committed a crime or other inappropriate behavior. In this context,
punishment involves not only the hope that the behavior will cease, but
also elements of retribution or retaliation; part of the intent is to hurt the
person who has committed the crime. Seen as something that a wrongdoer
deserves, punishment has moral or ethical connotations. Authority figures such
as governments, police, churches, or parents impose punishment to inhibit inap-
propriate behavior—that is, to keep people from breaking laws or rules. Punish-
ment may involve prison time, a death sentence, fines, the threat of going to
hell, spanking, or scolding. However, the everyday meaning of punishment is
quite different from the technical definition of punishment used in behavior
modification.

People who are unfamiliar with the technical definition of punishment may
believe that the use of punishment in behavior modification is wrong or danger-
ous. It is unfortunate that Skinner adopted the term punishment, a term that has
an existing meaning and many negative connotations. As a student, it is important
for you to understand the technical definition of punishment in behavior modifica-
tion and to realize that it is very different from the common view of punishment
in society.

104 Chapter 6

On Terms: Punish Behavior, not People
â–  It is correct to say that you punish a behavior (or a response). You are weakening a behavior by
punishing it. To say “The teacher punished Sarah’s disruptive behavior with time out” is correct.

■ It is incorrect to say that you punish a person. You don’t weaken a person, you weaken a person’s
behavior. To say, “The teacher punished Sarah for disruptive behavior” is not correct.

Positive and Negative Punishment

The two basic procedural variations of punishment are positive punishment and
negative punishment. The difference between positive and negative punishment
is determined by the consequence of the behavior. Positive punishment is
defined as follows.

1. The occurrence of a behavior
2. is followed by the presentation of an aversive stimulus,
3. and as a result, the behavior is less likely to occur in the future.

Negative punishment is defined as follows.

1. The occurrence of a behavior
2. is followed by the removal of a reinforcing stimulus,
3. and as a result, the behavior is less likely to occur in the future.

Notice that these definitions parallel the definitions of positive and
negative reinforcement (see Chapter 4). The critical difference is that reinforce-
ment strengthens a behavior or makes it more likely to occur in the future, whereas
punishment weakens a behavior or makes it less likely to occur in the future.

Many researchers have examined the effects of punishment on the behavior
of laboratory animals. Azrin and Holz (1966) discussed the early animal research
on punishment, much of which they had conducted themselves. Since then,
researchers have investigated the effects of positive and negative punishment on
human behavior (Axelrod & Apsche, 1983). For example, Corte, Wolf, and Locke
(1971) helped institutionalized adolescents with intellectual disabilities decrease
self-injurious behavior by using punishment. One subject slapped herself in the
face. Each time she did so, the researchers immediately applied a brief electric
shock with a handheld shock device. (Although the shock was painful, it did not
harm the girl.) As a result of this procedure, the number of times she slapped herself
in the face each hour decreased immediately from 300–400 to almost zero. (Note
that this study is from 1971. Electric shock is rarely, if ever, used as a punisher
today because of ethical concerns. This study is cited to illustrate the basic principle
of positive punishment, not to support the use of electric shock as a punisher.)

Why is this an example of positive punishment?

This is an example of positive punishment because the painful stimulus was
presented each time the girl slapped her face, and the behavior decreased as a

Punishment 105

result. Sajwaj, Libet, and Agras (1974) also used positive punishment to decrease
life-threatening rumination behavior in a 6-month-old infant. Rumination in
infants involves repeatedly regurgitating food into the mouth and swallowing it
again. It can result in dehydration, malnutrition, and even death. In this study,
each time the infant engaged in rumination, the researchers squirted a small
amount of lemon juice into her mouth. As a result, the rumination behavior
immediately decreased, and the infant began to gain weight.

One other form of positive punishment is based on the Premack principle,
which states that when a person is made to engage in a low-probability behavior
contingent on a high-probability behavior, the high-probability behavior will
decrease in frequency (Miltenberger & Fuqua, 1981). That is, if, after engaging
in a problem behavior, a person has to do something he or she doesn’t want to
do, the person will be less likely to engage in the problem behavior in the future.
Luce, Delquadri, and Hall (1980) used this principle to help a developmentally
delayed 6-year-old boy stop engaging in aggressive behavior. Each time the boy
hit someone in the classroom, he was required to stand up and sit down on the
floor ten times in a row. As shown in Figure 6-1, this punishment procedure,
called contingent exercise, resulted in an immediate decrease in the hitting
behavior.

One thing you should notice in Figure 6-1 is that punishment results in an
immediate decrease in the target behavior. Although extinction also decreases a

N
um

be
r o

f H
its

5 10 15 20
0

10

20

30

40

50

60

70

25 30 35 40

Baseline Baseline Contingent ExerciseContingent Exercise

•••
••

•

• •

•

•

•

•• •
•

••

••••••• •• ••
•

••

•• • •• •• •• •

FIGURE 6-1 In this graph, a positive punishment procedure called contingent exercise reduced the aggressive
behavior of a 6-year-old boy. This is an A-B-A-B research design, in which the baseline and treat-
ment conditions are implemented twice. (From Luce, S., Delquadri, J., & Hall, R. V. [1980]. Contin-
gent exercise: A mild but powerful procedure for suppressing inappropriate verbal and aggressive
behavior. Journal of Applied Behavior Analysis, 13, 583–594. Copyright © 1980 University of
Kansas Press. Reprinted by permission of the publisher.)

106 Chapter 6

behavior, it usually takes longer for the behavior to decrease, and an extinction
burst often occurs where the behavior increases briefly before it decreases. With
punishment, the decrease in behavior typically is immediate and there is no
extinction burst. However, other side effects are associated with the use of punish-
ment; these are described later in this chapter.

Negative punishment has also been the subject of extensive research. Two
examples of negative punishment are time-out from positive reinforcement and
response cost (see Chapter 17 for a more detailed discussion). Both involve the
loss of a reinforcing stimulus or activity after the occurrence of a problem
behavior. Some students may confuse negative punishment and extinction.
They both weaken behavior. Extinction involves withholding the reinforcer that
was maintaining the behavior. Negative punishment, by contrast, involves
removing or withdrawing a positive reinforcer after the behavior; the reinforcer
that is removed in negative punishment is one the individual had already
acquired and is not necessarily the same reinforcer that was maintaining the
behavior. For example, Johnny interrupts his parents and the behavior is rein-
forced by his parents’ attention. (They scold him each time he interrupts.) In
this case, extinction would involve withholding the parents’ attention each time
Johnny interrupts. Negative punishment would involve the loss of some other
reinforcer—such as allowance money or the opportunity to watch TV—each
time he interrupted. Both procedures would result in a decrease in the fre-
quency of interrupting.

Clark, Rowbury, Baer, and Baer (1973) used time-out to decrease aggressive
and disruptive behavior in an 8-year-old girl with Down syndrome. In time-out,
the person is removed from a reinforcing situation for a brief period after the prob-
lem behavior occurs. Each time the girl engaged in the problem behavior in the
classroom, she had to sit by herself in a small time-out room for 3 minutes. As a
result of time-out, her problem behaviors decreased immediately (Figure 6-2).
Through the use of time-out, the problem behavior was followed by the loss of
access to attention (social reinforcement) from the teacher and other reinforcers
in the classroom (Figure 6-3).

In a study by Phillips, Phillips, Fixsen, and Wolf (1971), “predelinquent”
youths with serious behavior problems in a residential treatment program earned
points for engaging in appropriate behavior and traded in their points for backup
reinforcers such as snacks, money, and privileges. The points were conditioned
reinforcers. The researchers then used a negative punishment procedure called
response cost to decrease late arrivals for supper. When the youths arrived late,
they lost some of the points they had earned. As a result, late arrivals decreased
until the youths always showed up on time.

Positive punishment and negative punishment sometimes are called other
names that are more descriptive. However, it is simpler to speak of positive punish-
ment and negative punishment, and these terms are parallel with positive rein-
forcement and negative reinforcement.

Look at the examples of punishment in Table 6-1. Which are examples of posi-
tive punishment and which are examples of negative punishment? Answers are
provided at the end of the chapter in Appendix A.

Punishment 107

In all of these examples, the process resulted in a decrease in the future occur-
rence of the behavior. Therefore, in each example, the presentation or removal of
a stimulus as a consequence of the behavior functioned as punishment.

Baseline Time-Out

Chokes and ArmwrapsChokes and Armwraps

Baseline Time-Out

Other Attacks toward PeopleOther Attacks toward People

N
um

be
r o

f 1
0-

Se
co

nd
In

te
rv

al
s

pe
r H

ou
r C

on
ta

in
in

g
D

is
ru

pt
iv

e
Be

ha
vi

or

40

50

60

20

30

0

10

70

1 9 17 27

Baseline Time-Out

Attacks toward MaterialsAttacks toward Materials

•
••••

•

•

•
•

•
••• •• • •• ••••••

••

• ••

••
••

•

••

•
•

•
•

•

•

••

•

•

•

•••• • ••

••

• •
•

•
••• ••

•

•

•

•

•

•

•

••••••
• ••

Days

20

30

0

10

20

30

0

10

FIGURE 6-2 The effect of a negative punishment procedure (time-out) on the aggressive and disruptive
behavior of a young girl with Down syndrome is shown. This graph illustrates a multiple-
baseline-across-behaviors design. Time-out was implemented for three different behaviors of one
subject, and the use of time-out was staggered over time. (From Clark, H., Rowbury, T., Baer, A., &
Baer, D. [1973]. Time out as a punishing stimulus in continuous and intermittent schedules. Jour-
nal of Applied Behavior Analysis, 6, 443–455. Reprinted by permission of the author.)

108 Chapter 6

Other Names for Positive Punishment
â–  Punishment by application

â–  Punishment by contingent presentation of a stimulus

â–  Punishment by presentation of an aversive stimulus

â–  Response-contingent presentation of a punisher

Other Names for Negative Punishment
â–  Punishment by withdrawal

â–  Punishment by loss of reinforcers

â–  The penalty contingency

â–  Response-contingent removal of a positive reinforcer

Sources: Cooper, Heron, & Heward, 1987; Malott, Whaley, & Malott, 1993; Sundel & Sundel,
1993.

FIGURE 6-3 This 8-year-old child has to sit in the small time-out room by herself each time she engages in
aggressive behavior in the classroom. By sitting in the time-out room, she loses access to such
reinforcers as teacher attention, attention from other students, and toys. As a result, the aggressive
behavior decreases.

Punishment 109

On Terms: Distinguishing Between Positive and Negative Punishment

Some students have confusion distinguishing between positive and negative punishment. They are both
types of punishment, therefore, they both weaken behavior. The only difference is whether a stimulus is
added (positive punishment) or removed (negative punishment) following the behavior. Think of positive
as a plus or addition (þ) sign and negative as a minus or subtraction (�) sign. In þ punishment, you
add a stimulus (an aversive stimulus) after the behavior. In � punishment, you subtract or take away a
stimulus (a reinforcer) after the behavior. If you think of positive and negative in terms of adding or sub-
tracting a stimulus after the behavior, the distinction should be more clear.

Unconditioned and Conditioned Punishers

Like reinforcement, punishment is a natural process that affects human behavior. Some
events or stimuli are naturally punishing because avoiding or minimizing contact with
these stimuli has survival value (Cooper et al., 1987). Painful stimuli or extreme levels
of stimulation often are dangerous. Behaviors that produce painful or extreme stimula-
tion are naturally weakened, and behaviors that result in escape or avoidance of such
stimulation are naturally strengthened. For this reason, painful stimuli or extreme levels
of stimulation have biological importance. Such stimuli are called unconditioned
punishers. Through the process of evolution, humans have developed the capacity for
their behavior to be punished by these naturally aversive events without any prior train-
ing or experience. For example, extreme heat or cold, extreme levels of auditory or
visual stimulation, or any painful stimulus (e.g., from electric shock, a sharp object, or
a forceful blow) naturally weakens the behavior that produces it. If these were not
unconditioned punishers, we would be more likely to engage in dangerous behaviors
that could result in injury or death. We quickly learn not to put our hands into a fire,
look directly into the sun, touch sharp objects, or go barefoot in the snow or on hot
asphalt because each of these behaviors results in a naturally punishing consequence.

A second type of punishing stimulus is called a conditioned punisher. Condi-
tioned punishers are stimuli or events that function as punishers only after being paired
with unconditioned punishers or other existing conditioned punishers. Any stimulus or
event may become a conditioned punisher if it is paired with an established punisher.

The word no is a common conditioned punisher. Because it is often paired
with many other punishing stimuli, it eventually becomes a punisher itself. For
example, if a child reaches for an electrical outlet and the parent says “no,” the
child may be less likely to reach for the outlet in the future. When the child spells
a word incorrectly in the classroom and the teacher says “no,” the child will be
less likely to spell that word incorrectly in the future. The word no is considered
a generalized conditioned punisher because it has been paired with a variety of
other unconditioned and conditioned punishers over the course of a person’s life.
Van Houten and his colleagues (Van Houten, Nau, MacKenzie-Keating,
Sameoto, & Colavecchia, 1982) found that if firm reprimands were delivered to
students in the classroom when they engaged in disruptive behavior, their disrup-
tive behavior decreased. In this study, reprimands were conditioned punishers for
the students’ disruptive behavior. Threats of harm often are conditioned punishers.
Because threats have often been associated with painful stimulation in the past,
threats may become conditioned punishers.

110 Chapter 6

Stimuli that are associated with the loss of reinforcers may become condi-
tioned punishers. A parking ticket or a speeding ticket is associated with the loss
of money (paying a fine), so the ticket is a conditioned punisher for many people.
In reality, whether speeding tickets or parking tickets function as conditioned pun-
ishers depends on a number of factors, including the schedule of punishment and
the magnitude of the punishing stimulus. These and other factors that influence
the effectiveness of punishment are discussed later in this chapter.

A warning from a parent may become a conditioned punisher if it has been
paired with the loss of reinforcers such as allowance money, privileges, or preferred
activities. As a result, when a child misbehaves and the parent gives the child a
warning, the child may be less likely to engage in the same misbehavior in the
future. A facial expression or look of disapproval may be a conditioned punisher
when it is associated with the loss of attention or approval from an important person
(such as a parent or teacher). A facial expression may also be associated with an aver-
sive event such as a scolding or a spanking, and thus may function as a conditioned
punisher (Doleys, Wells, Hobbs, Roberts, & Cartelli, 1976; Jones & Miller, 1974).

Once again, it is important to remember that a conditioned punisher is
defined functionally. It is defined as a punisher only if it weakens the behavior
that it follows. If a person exceeds the speed limit and receives a speeding ticket
and the outcome is that the person is less likely to speed in the future, the ticket
functioned as a punisher. However, if the person continues to speed after
receiving a ticket, the ticket was not a punisher. Consider the following
example.

Is the mother’s angry look a conditioned punisher in this situation? Why or why
not?

The look is not a conditioned punisher because the child’s behavior of belch-
ing at the table was not weakened; the child did not stop engaging in the behav-
ior. The mother’s look may have functioned as a positive reinforcer, or perhaps
other family members laughed when the child belched, and thus reinforced the
belching behavior. Alternatively, belching may be naturally reinforcing because it
relieves an unpleasant sensation in the stomach.

Contrasting Reinforcement and Punishment

Important similarities and differences exist between positive and negative rein-
forcement on one hand and positive and negative punishment on the other. The
defining features of each principle are that a behavior is followed by a conse-
quence, and the consequence influences the future occurrence of the behavior.

Response

Outcome: Child continues to belch at the dinner table in the future.

Child belches at the dinner table and immediately mom gives the child an angry look.

Punishment 111

The similarities and differences between the two types of reinforcement and pun-
ishment can be summarized as follows:

Note that, when a stimulus is presented after a behavior (left column), the process may
be positive reinforcement or positive punishment, depending on whether the
behavior is strengthened (reinforcement) or weakened (punishment) in the future.
When a stimulus is removed after the behavior (right column), the process may be
negative reinforcement or negative punishment. It is negative reinforcement if the
behavior is strengthened and negative punishment if the behavior is weakened. Note
that when a behavior is strengthened, the process is reinforcement (positive or
negative). Also note that when a behavior is weakened, the process is punishment
(positive or negative).

One particular stimulus may be involved in reinforcement and punishment of
different behaviors in the same situation, depending on whether the stimulus is
presented or removed after the behavior. Consider the example of Kathy and the
dog. When Kathy reached over the fence, this behavior was followed immediately
by the presentation of an aversive stimulus (the dog bit her). The dog’s bite served
as a punisher: Kathy was less likely to reach over the fence in the future. However,
when Kathy pulled her hand back quickly, she terminated the dog bite. Because
pulling her hand back removed the pain of being bitten, this behavior was

Behavior is strengthened (increases in the future)

Behavior is weakened (decreases in the future)

Outcome

Positive reinforcement

Positive punishment

Negative reinforcement

Negative punishment

Stimulus is presented

Consequence of the Behavior

Stimulus is removed

Response

Outcome: Kathy is less likely to reach over the fence in the future.

Kathy reached over the fence and immediately the dog bit her.

Positive Punishment

Response

Outcome: Kathy is more likely to pull her hand back when presented with a similar painful stimulus.

Kathy pulled her hand back and immediately she terminated the dog bite.

Negative Reinforcement

112 Chapter 6

strengthened. This is an example of negative reinforcement. As you can see, when
the dog bite was presented after one behavior, the behavior was weakened; when
the dog bite was removed after another behavior, that behavior was strengthened.

In the example of Otis and the hot skillet, the immediate consequence of
grabbing the skillet handle was a painful stimulus. The outcome was that Otis
was less likely to grab a hot skillet in the future. This is positive punishment.

How is negative reinforcement involved in this example?

When Otis used a hot pad, he avoided the painful stimulus. As a result, he
is more likely to use a hot pad when grabbing a hot skillet in the future (nega-
tive reinforcement). Touching the hot skillet is punished by the presentation of
a painful stimulus; using the hot pad is reinforced by avoidance of the painful
stimulus.

Now consider how the same stimulus may be involved in negative punish-
ment of one behavior and positive reinforcement of another behavior. If a rein-
forcing stimulus is removed after a behavior, the behavior will decrease in the
future (negative punishment), but if a reinforcing stimulus is presented after a
behavior, the behavior will increase in the future (positive reinforcement). You
know that a stimulus is functioning as a positive reinforcer when its presentation
after a behavior increases that behavior and its removal after a behavior decreases
that behavior. For example, Fred’s parents take his bicycle away for a week
whenever they catch him riding after dark. This makes Fred less likely to ride
his bike after dark (negative punishment). However, after a few days, Fred pleads
with his parents to let him ride his bike again and promises never to ride after
dark. They give in and give him his bike back. As a result, he is more likely to
plead with his parents in the future when his bike is taken away (positive
reinforcement).

Response

Outcome: Fred is less likely to ride his bike after dark.

Fred rides his bike after dark and then bike is removed for 1 week.

Negative Punishment

Response

Outcome: Fred is more likely to plead with his parents when he doesn’t have his bike.

Fred pleads with his parents and then bike is presented to Fred.

Positive Reinforcement

Punishment 113

Factors That Influence the Effectiveness
of Punishment

The factors that influence the effectiveness of punishment are similar to those that
influence reinforcement. They include immediacy, contingency, establishing
operations, individual differences, and magnitude.

Immediacy
When a punishing stimulus immediately follows a behavior, or when the loss of a
reinforcer occurs immediately after the behavior, the behavior is more likely to be
weakened. That is, for punishment to be most effective, the consequence must
follow the behavior immediately. As the delay between the behavior and the con-
sequence increases, the effectiveness of the consequence as a punisher decreases.

To illustrate this point, consider what would happen if a punishing stimulus
occurred some time after the behavior occurred. A student makes a sarcastic comment
in class and the teacher immediately gives her an angry look. As a result, the student is
less likely to make a sarcastic comment in class. If the teacher had given the student an
angry look 30 minutes after the student made the sarcastic comment, the look would
not function as a punisher for the behavior of making sarcastic comments. Instead,
the teacher’s angry look probably would have functioned as a punisher for whatever
behavior the student had engaged in immediately before the look.

Contingency
For punishment to be most effective, the punishing stimulus should occur every time
the behavior occurs. We would say that the punishing consequence is contingent on
the behavior when the punisher follows the behavior each time the behavior occurs
and the punisher does not occur when the behavior does not occur. A punisher is
most likely to weaken a behavior when it is contingent on the behavior. This means
that punishment is less effective when it is applied inconsistently—that is, when the
punisher follows only some occurrences of the behavior or when the punisher is pre-
sented in the absence of the behavior. If a reinforcement schedule continues to be in
effect for the behavior, and punishment is applied inconsistently, some occurrences of
the behavior may be followed by a punisher and some occurrences of the behavior
may be followed by a reinforcer. In this case, the behavior is being influenced by an
intermittent schedule of reinforcement at the same time that it is resulting in an inter-
mittent punishment schedule. When a concurrent schedule of reinforcement is com-
peting with punishment, the effects of punishment are likely to be diminished.

If a hungry rat presses a bar in an experimental chamber and receives food pellets,
the rat will continue to press the bar. However, if punishment is implemented and the
rat receives an electric shock each time it presses the bar, the bar-pressing behavior will
stop. Now suppose that the rat continues to receive food for pressing the bar and
receives a shock only occasionally when it presses the bar. In this case, the punishing
stimulus would not be effective because it is applied inconsistently or intermittently.
The effect of the punishing stimulus in this case depends on the magnitude of the
stimulus (how strong the shock is), how often it follows the behavior, and the magni-
tude of the establishing operation for food (how hungry the rat is).

114 Chapter 6

Motivating Operations
Just as establishing operations (EOs) and abolishing operations (AOs) may influ-
ence the effectiveness of reinforcers, they also influence the effectiveness of pun-
ishers. An establishing operation is an event or a condition that makes a
consequence more effective as a punisher (or a reinforcer). An abolishing opera-
tion is an event or a condition that makes a consequence less effective as a pun-
isher (or a reinforcer). In the case of negative punishment, deprivation is an EO
makes the loss of reinforcers more effective as a punisher and satiation is an AO
that makes the loss of reinforcers less effective as a punisher. For example, telling
a child who misbehaves at the dinner table that dessert will be taken away will: a)
be a more effective punisher if the child has not eaten any dessert yet and is still
hungry (EO), b) be a less effective punisher if the child has had two or three help-
ings of the dessert already and is no longer hungry (AO). Losing allowance money
for misbehavior will: a) be a more effective punisher if the child has no other
money and plans to buy a toy with the allowance money (EO), b) be a less effec-
tive punisher if the child has recently received money from other sources (AO).

In the case of positive punishment, any event or condition that enhances
the aversiveness of a stimulus event makes that event a more effective punisher
(EO), whereas events that minimize the aversiveness of a stimulus event make it
less effective as a punisher (AO). For example, some drugs (e.g., morphine) min-
imize the effectiveness of a painful stimulus as a punisher. Other drugs (e.g.,
alcohol) may reduce the effectiveness of social stimuli (e.g., peer disapproval) as
punishers.

Are these examples of AOs or EOs?
These are examples of AOs because in each case the drugs made punishers less
effective. Instructions or rules may enhance the effectiveness of certain stimuli as
punishers. For example, a carpenter tells his apprentice that when the electric
saw starts to vibrate, it may damage the saw or break the blade. As a result of this
instruction, vibration from the electric saw is established as a punisher. The behav-
ior that produces the vibration (e.g., sawing at an angle, pushing too hard on the
saw) is weakened.

Is this an example of an EO or an AO?
This is an example of an EO because the instruction made the presence of vibra-
tion more aversive or more effective as a punisher for using the saw incorrectly. In
addition, using the saw correctly avoids the vibration and this behavior is strength-
ened through negative reinforcement.

Factors That Influence the Effectiveness of Punishment
Immediacy A stimulus is more effective as a punisher when presented

immediately after the behavior.

Contingency A stimulus is more effective as a punisher when presented
contingent on the behavior.

(continued)

Punishment 115

Motivating operations Some antecedent events make a stimulus more effective
as a punisher at a particular time (EO). Some events
make a stimulus a less effective punisher at a particular
time (AO).

Individual differences and magnitude Punishers vary from person to person. In general, a
more intense aversive stimulus is a more effective
punisher.

Individual Differences and Magnitude of the Punisher
Another factor that influences the effectiveness of punishment is the nature of the
punishing consequence. The events that function as punishers vary from person to
person (Fisher et al., 1994). Some events may be established as conditioned punish-
ers for some people and not for others because people have different experiences or
conditioning histories. Likewise, whether a stimulus functions as a punisher depends
on its magnitude or intensity. In general, a more intense aversive stimulus is more
likely to function as a punisher. This also varies from person to person. For example,
a mosquito bite is a mildly aversive stimulus for most people; thus, the behavior of
wearing shorts in the woods may be punished by mosquito bites on the legs, and
wearing long pants may be negatively reinforced by the avoidance of mosquito
bites. However, some people refuse to go outside at all when the mosquitoes are bit-
ing, whereas others go outside and do not seem to be bothered by mosquito bites.
This suggests that mosquito bites may be a punishing stimulus for some people but
not others. The more intense pain of a bee sting, by contrast, probably is a punisher
for most people. People will stop engaging in the behavior that resulted in a bee
sting and will engage in other behaviors to avoid a bee sting. Because the bee sting
is more intense than a mosquito bite, it is more likely to be an effective punisher.

FOR FURTHER READING
Factors That Influence Punishment

The behavior modification principle of punishment has been studied by researchers for years. One
important recommendation when using punishment is to use a reinforcement procedure in conjunction
with punishment. For example, Thompson, Iwata, Conners, and Roscoe (1999) showed that punish-
ment for self-injurious behavior was more effective when a differential reinforcement procedure was
used with punishment (they reinforced a desirable behavior at the same time they used punishment).
Similarly, Hanley, Piazza, Fisher, and Maglieri (2005) showed that when punishment was added to a
differential reinforcement procedure, the reinforcement procedure was more effective. Interestingly,
the children in this study preferred the procedure involving reinforcement and punishment over rein-
forcement alone. These two studies demonstrate the importance of combining reinforcement and pun-
ishment. In an investigation of different intensities of punishment, Vorndan and Lerman (2006)
showed that a less intrusive punishment procedure was not effective until it was paired with a more
intrusive punishment procedure. Finally, Lerman, Iwata, Shore, and DeLeon (1997) showed that inter-
mittent punishment is less effective than continuous punishment, although for some participants,
intermittent punishment was effective when it followed the use of continuous punishment. Together,
these two studies suggest that the punishment contingency and intensity are important factors in the
effectiveness of punishment.

Factors That Influence the Effectiveness of Punishment (Continued )

116 Chapter 6

Problems with Punishment

A number of problems or issues must be considered with the use of punishment,
especially positive punishment involving the use of painful or other aversive stimuli.

â–  Punishment may produce elicited aggression or other emotional side
effects.

â–  The use of punishment may result in escape or avoidance behaviors by the
person whose behavior is being punished.

â–  The use of punishment may be negatively reinforcing for the person using
punishment, and thus may result in the misuse or overuse of punishment.

â–  When punishment is used, its use is modeled, and observers or people
whose behavior is punished may be more likely to use punishment them-
selves in the future.

â–  Finally, punishment is associated with a number of ethical issues and issues
of acceptability. These issues are addressed in detail in Chapter 18.

Emotional Reactions to Punishment
Behavioral research with nonhuman subjects has demonstrated that aggressive
behavior and other emotional responses may occur when painful stimuli are pre-
sented as punishers. For example, Azrin, Hutchinson, and Hake (1963) showed
that presenting a painful stimulus (shock) results in aggressive behavior in labora-
tory animals. In this study, when one monkey received a shock, it immediately
attacked another monkey that was present when the shock was delivered. When
such aggressive behaviors or other emotional responses result in the termination
of the painful or aversive stimulus, they are negatively reinforced. Thus, the ten-
dency to engage in aggressive behavior (especially when it is directed at the source
of the aversive stimulus) may have survival value.

Escape and Avoidance
Whenever an aversive stimulus is used in a punishment procedure, an opportunity
for escape and avoidance behavior is created. Any behavior that functions to avoid
or escape from the presentation of an aversive stimulus is strengthened through
negative reinforcement. Therefore, although an aversive stimulus may be pre-
sented after a target behavior to decrease the target behavior, any behavior the per-
son engages in to terminate or avoid that aversive stimulus is reinforced (Azrin,
Hake, Holz, & Hutchinson, 1965). For example, a child might run away or hide
from a parent who is about to spank the child. Sometimes people learn to lie to
avoid punishment, or learn to avoid the person who delivers the punishing stimu-
lus. When implementing a punishment procedure, you have to be careful that
inappropriate escape and avoidance behaviors do not develop.

Negative Reinforcement for the Use of Punishment
Some authors argue that punishment may be too easily misused or overused
because its use is negatively reinforcing to the person implementing it (Sulzer-
Azaroff & Mayer, 1991).

Punishment 117

Describe how the use of punishment may be negatively reinforcing.
When punishment is used, it results in an immediate decrease in the problem
behavior. If the behavior decreased by punishment is aversive to the person using
punishment, the use of punishment is negatively reinforced by the termination of
the aversive behavior. As a result, the person is more likely to use punishment in
the future in similar circumstances. For example, Dr. Hopkins hated it when her
students talked in class while she was teaching. Whenever someone talked in
class, Dr. Hopkins stopped teaching and stared at the student with her meanest
look. When she did this, the student immediately stopped talking in class. As a
result, Dr. Hopkins’s behavior of staring at students was reinforced by the termina-
tion of the students’ talking in class. Dr. Hopkins used the stare frequently, and
she was known all over the university for it.

Punishment and Modeling
People who observe someone making frequent use of punishment may themselves
be more likely to use punishment when they are in similar situations. This is espe-
cially true with children, for whom observational learning plays a major role in
the development of appropriate and inappropriate behaviors (Figure 6-4). For
example, children who experience frequent spanking or observe aggressive behav-
ior may be more likely to engage in aggressive behavior themselves (Bandura,
1969; Bandura, Ross, & Ross, 1963).

FIGURE 6-4 One of the possible problems with punishment is observational learning, as illustrated here. To
punish her daughter’s misbehavior, a mother uses spanking. As a result of observing her mother,
the child also engages in the behavior with her doll.

118 Chapter 6

Ethical Issues
Some debate exists among professionals about whether it is ethical to use punish-
ment, especially painful or aversive stimuli, to change the behavior of others
(Repp & Singh, 1990). Some argue that the use of punishment cannot be justified
(Meyer & Evans, 1989). Others argue that the use of punishment may be justified
if the behavior is harmful or serious enough and, therefore, the potential benefits
to the individual are great (Linscheid, Iwata, Ricketts, Williams, & Griffin, 1990).
Clearly, ethical issues must be considered before punishment is used as a behavior
modification procedure. Surveys show that procedures involving punishment are
much less acceptable in the profession than are behavior modification procedures
that use reinforcement or other principles (Kazdin, 1980; Miltenberger, Lennox, &
Erfanian, 1989). Professionals must consider a number of issues before they
decide to use behavior modification procedures based on punishment. In addi-
tion, punishment procedures are always used in conjunction with positive rein-
forcement procedures to strengthen the desirable behavior. (See Chapter 18 for
further discussion of these issues.)

CHAPTER SUMMARY

1. Punishment is a basic principle of behavior. Its
definition has three basic components: The
occurrence of a behavior is followed by an imme-
diate consequence, and the behavior is less likely
to occur in the future.

2. A common misconception about punishment is
that it means doing harm to another person or
exacting retribution on another person for that
person’s misbehavior. Instead, punishment is a
label for a behavioral principle devoid of the
legal or moral connotations usually associated
with the word.

3. There are two procedural variations of punish-
ment: positive and negative punishment. In posi-
tive punishment, an aversive stimulus is presented
after the behavior. In negative punishment, a rein-
forcing stimulus is removed after the behavior. In

both cases, the behavior is less likely to occur in
the future.

4. The two types of punishing stimuli are uncondi-
tioned punishers and conditioned punishers. An
unconditioned punisher is naturally punishing.
A conditioned punisher is developed by pairing a
neutral stimulus with an unconditioned punisher
or another conditioned punisher.

5. Factors that influence the effectiveness of punish-
ment include immediacy, contingency, motivating
operations, individual differences, and magnitude.

6. Potential problems associated with the use of pun-
ishment include emotional reactions to punish-
ment, the development of escape and avoidance
behaviors, negative reinforcement for the use of
punishment, modeling of the use of punishment,
and ethical issues.

KEY TERMS

conditioned punisher, 110
generalized conditioned

punisher, 110
negative punishment, 105

positive punishment, 105
punisher, 102
punishment, 102
response cost, 107

time-out from positive
reinforcement, 107

unconditioned punisher, 110

Punishment 119

PRACTICE TEST

1. Define punishment. (p. 102)
2. In ordinary usage, what does punishment mean?

How does this contrast with the definition of
punishment in behavior modification? (p. 104)

3. Provide an example of punishment from your
own life. (b) Is this an example of positive or
negative punishment? Why? (c) Does this exam-
ple involve an unconditioned or a conditioned
punisher? Why?

4. The behavior modification definition of punish-
ment is a functional definition. What do we
mean by functional definition? (p. 103)

5. Define positive punishment. Provide an exam-
ple. What other terms are sometimes used in
place of positive punishment? (pp. 105–109)

6. Define negative punishment. Provide an exam-
ple. What other terms are sometimes used in
place of negative punishment? (pp. 105–109)

7. (a) What is an unconditioned punisher? (b) What
does it mean to say that a punishing stimulus has
biological importance? (c) Provide some exam-
ples of unconditioned punishers. (p. 110)

8. (a) What is a conditioned punisher? (b) How is a
neutral stimulus established as a conditioned pun-
isher? (c) Provide some examples of conditioned
punishers from your own life. (pp. 110–111)

9. Describe how a painful stimulus may be
involved in both positive punishment and nega-
tive reinforcement. Provide an example. (p. 112)

10. Describe how a reinforcing stimulus may be
involved in both negative punishment and posi-
tive reinforcement. Provide an example. (p. 113)

11. Describe how immediacy influences the effec-
tiveness of punishment. (p. 114)

12. How does consistency or the schedule of punish-
ment influence the effectiveness of punishment?
(p. 114)

13. What is an establishing operation? Provide an
example of an establishing operation that influ-
ences the effectiveness of punishment. What is
an abolishing operation? Provide an example of
an abolishing operation that influences the
effectiveness of punishment. (pp. 115)

14. How is the intensity of a stimulus related to its
effectiveness as a punisher? (pp. 116)

15. Describe five problems that may be associated
with the use of punishment. (pp. 117–119)

16. Identify each of the following as an example of
positive punishment, negative punishment, or
extinction. When analyzing each example, be
sure to ask yourself three questions: (pp. 105–109)
â–  What is the behavior?
â–  What happened immediately after the behav-
ior? (Was a stimulus added or removed, or was
the reinforcer for the behavior terminated?)

â–  What happened to the behavior in the future?
(Was the behavior weakened? Is it less likely to
occur?)
a. Rachel got up early every morning and

raided the cookie jar. Her mom realized
what was going on and stopped putting
cookies in the jar. After this, when Rachel
reached in the cookie jar she no longer
found cookies. As a result, she no longer
raided the cookie jar.

b. Heather tossed eggs at the school during
Halloween. The principal caught her and
made her wash all the windows in the
school. Heather never threw eggs at the
school again.

c. Doug threw eggs at his neighbors’ house
during Halloween. His parents caught him
and made him give his neighbors $100 to
get their house cleaned. Doug never threw
eggs at the neighbors’ house again.

d. Ralph acted out in class and his teacher
gave him a mean look. After this, Ralph
never acted out in class again.

e. Suzie watched a lot of television and used
the remote control to turn it on and to
change channels. One day the remote did
not work. She tried it a few times and even-
tually quit using it.

f. Bill hit his sister and his mom took his
allowance away for that week. As a result,
he doesn’t hit his sister anymore.

g. Amanda tried to climb the fence into an
apple orchard. The fence was electrified
and gave her a shock. As a result, she
doesn’t climb that fence anymore.

120 Chapter 6

APPENDIX A
Examples of Positive Punishment and Negative Punishment from Table 6-1

1. Positive punishment. The behavior of looking
down while riding resulted in the presentation
of a painful stimulus when Ed hit the car.

2. Negative punishment. The behavior of hitting
resulted in the removal of the opportunity to
play with her toys and her friends.

3. Negative punishment. Running the lawn mower
over the hose resulted in the loss of money.

4. Positivepunishment.Readingwhiledrivingwasimme-
diately followed by the occurrence of an accident.

5. Negative punishment. Each time Helen got out of
her seat, the consequence was the removal of a
poker chip.

6. Positive punishment. Kevin’s telling jokes about
his wife’s cooking resulted in the presentation of
an aversive stimulus: an icy stare from his wife.

Punishment 121

This page intentionally left blank

Chapter Seven

Stimulus Control: Discrimination
and Generalization

In discussing reinforcement, extinction, and punishment, we saw the importanceof consequences in the control of operant behavior. Operant behavior is
strengthened when it is followed by a reinforcing consequence; it is weakened
when the reinforcing consequence no longer follows the behavior (extinction).
A punishing consequence also weakens the behavior. These basic principles of
behavior—reinforcement, extinction, and punishment—explain why behaviors

increase and continue to occur or decrease and stop occurring.
Because operant behavior is controlled by its consequences, behav-
ior analysts analyze the events that follow the behavior to under-
stand why it is occurring, and they manipulate the consequences of
the behavior to modify it.

This chapter expands the analysis of operant behavior and dis-
cusses the importance of antecedents, stimulus events that precede
an operant response. The antecedents of a behavior are the stimuli,
events, situations, or circumstances that are present when it occurs
or were present immediately before the behavior. To understand
and modify operant behavior, it is important to analyze the antece-
dents, as well as the consequences of the behavior. Therefore, this
chapter focuses on antecedents, behavior, and consequences, the
ABCs of operant behavior.

Why is it important to understand the antecedents of operant behavior?

When we understand the antecedents of operant behavior, we have informa-
tion on the circumstances in which the behavior was reinforced and the circum-
stances in which the behavior was not reinforced or was punished. A behavior
continues to occur in situations in which it has been reinforced in the past, and
stops occurring in situations in which it has not been reinforced or has been
punished in the past. As you can see, the effects of reinforcement, extinction,
and punishment are situation-specific. Consider the following examples.

â–  What is an antecedent stimulus and
how is it involved in stimulus control of
operant behavior?

â–  How is stimulus control developed
through stimulus discrimination
training?

â–  What is the three-term contingency?

â–  What is generalization and how does it
differ from discrimination?

123

Examples of Stimulus Control

Whenever Jake wants some extra cash to spend, he asks his mom and she usually
gives him some money. When he asks his dad, his dad usually refuses to give him
any money and tells him to get a job. As a result, he usually asks his mom for
money instead of his dad.

As you can see, the behavior of asking for money was reinforced in one situation
(with his mom) but was not reinforced in another situation (with his dad). Therefore,
the behavior continues to occur in the situation in which it was reinforced and no lon-
ger occurs in the situation in which it was not reinforced: Jake asks only his mom for
money. His mom’s presence is an antecedent for Jake’s behavior of asking for cash.
We would say that his mom’s presence has stimulus control over Jake’s behavior of ask-
ing for money. It is also important to note that Jake only asks his mom for money when
he needs it; that is, when an EO is present. If there is no EO (if he has nothing to buy),
he will not ask his mom for money.

Consider another example. Ginny decides she will go out back and pick a
few strawberries from the bushes in her backyard. When she picks a bright red
strawberry, it is sweet and juicy and tastes great. When she picks one that is still
slightly green, however, it is sour and hard and doesn’t taste very good. As she con-
tinues to pick the strawberries and eat them, she chooses only the red ones. A red
strawberry is an antecedent stimulus. The behavior of picking and eating a red
strawberry is reinforced. Therefore, she is more likely to pick and eat red ones.
The behavior of eating a green strawberry is not reinforced; she no longer picks
green ones. Picking and eating only red strawberries and not green ones is an
example of stimulus control. We would say that the presence of red strawberries
has stimulus control over Ginny’s behavior of picking and eating the strawberries.
It is important to note also that Ginny only picks red strawberries when an EO is
present (she is hungry, she needs strawberries for cooking, someone asks her to get
some strawberries, etc). If an EO is not present, she will not pick strawberries.

Antecedent

Outcome: Jake asks his mom for money in the future and does not ask his dad for money anymore.

Mom is present. Jake asks for money. Mom gives him the cash.

Consequence

Dad is present. Jake asks for money. Dad does not give him cash.

Behavior

Antecedent

Outcome: Ginny is likely to pick and eat red strawberries and to stop eating green ones.

Red strawberry Ginny picks and eats it. Tastes great.

Consequence

Green strawberry Ginny picks and eats it. Tastes awful.

Behavior

124 Chapter 7

Defining Stimulus Control

The two preceding examples illustrate the principle of stimulus control. In each,
a behavior was more likely to occur when a specific antecedent stimulus was pres-
ent. For Jake, the antecedent stimulus that was present when he asked for money
was his mom. For Ginny, the antecedent stimulus when she was picking and eat-
ing strawberries was the presence of red strawberries. A behavior is said to be
under stimulus control when there is an increased probability that the behavior
will occur in the presence of a specific antecedent stimulus or a stimulus from a
specific stimulus class. (Red strawberries are a stimulus class. Any one particular
red strawberry is a member of this stimulus class.)

What are some of your own behaviors that are under stimulus control?

To answer this question, ask yourself which of your behaviors occur only in
specific situations or in certain circumstances (i.e., when a specific antecedent
stimulus is present). What you will find is that almost all of your behaviors are
under stimulus control. Behaviors usually don’t occur randomly; they occur in
the specific situations or circumstances in which they were reinforced in the past.
Table 7-1 lists examples of some behaviors that are under stimulus control.

TABLE 7-1 Examples for Self-Assessment (Stimulus Control)

1. A man says “I love you” to his wife but not to any of the people where he works.

Antecedent�! Behavior�! Consequence
His wife is present. He says “I love you.” She says the same to him.

2. When the stop light is red, you stop. When it is green, you go.

Antecedent�! Behavior�! Consequence
Green light. You press the accelerator. You travel to where you are

going and avoid people honking
at you.

Antecedent�! Behavior�! Consequence
Red light. You press the brake pedal. You avoid an accident or a

traffic ticket.

3. You tell off-color jokes to your friends but not to your parents or teachers.

Antecedent�! Behavior�! Consequence
Your friends are present. You tell off-color jokes. They laugh and tell you jokes.

4. When the phone rings, you pick it up and talk to the person who called.

Antecedent�! Behavior�! Consequence
The phone rings. You answer the phone. You talk to the person who

called.

5. When the light on the rechargeable electric drill is on, you use the drill.

Antecedent�! Behavior�! Consequence
The light on the rechargeable
drill is on.

You take the drill and use it to
drill a hole.

The drill works fine.

Stimulus Control: Discrimination and Generalization 125

Each example in Table 7-1 shows an antecedent stimulus, a behavior, and a
consequence. In each example, the behavior is more likely to occur when the
antecedent stimulus is present. Why? The behavior occurs when the antecedent
is present because that is the only time the behavior has been reinforced.
Consider each example.

■ Saying “I love you” is reinforced by the man’s wife. If he said “I love you”
to people at work, they would not reinforce the behavior. (They might give him
strange looks or worse.) As a result, he says “I love you” only to his wife.

â–  Stopping at a red light is reinforced by avoiding an accident and a traffic
ticket (negative reinforcement). However, stopping at a green light would result
in people honking at you and making angry gestures (positive punishment).
Therefore, you stop at red lights and not at green lights.

â–  Telling an off-color joke to your friends is reinforced by laughs and atten-
tion. However, telling such jokes to your parents would not be reinforced and
may be punished with dirty looks or reprimands. Therefore, you tell off-color
jokes only to your friends.

â–  Picking up the phone when it rings is reinforced by talking to the caller;
picking the phone up when it does not ring is not reinforced because no one is
on the other end. As a result, you pick up the phone only when it rings (unless
you are making a call).

â–  When the charger light is on, using the drill is reinforced because the drill
works effectively. When the light is not on, using the drill is never reinforced
because the drill doesn’t work. As a result, you use the drill only when the light
is on.

Developing Stimulus Control: Stimulus
Discrimination Training

As you can see from the preceding examples, stimulus control develops because a
behavior is reinforced only in the presence of a particular antecedent stimulus.
Therefore, the behavior continues to occur in the future only when that anteced-
ent stimulus is present. The antecedent stimulus that is present when a behavior is
reinforced is known as the discriminative stimulus (SD). The process of reinfor-
cing a behavior only when a specific antecedent stimulus (SD) is present is called
stimulus discrimination training.

Two steps are involved in stimulus discrimination training.

1. When the SD is present, the behavior is reinforced.
2. When any other antecedent stimuli except the SD are present, the behav-

ior is not reinforced. During discrimination training, any antecedent stim-
ulus that is present when the behavior is not reinforced is called an
S-delta (S�).

As a result of discrimination training, a behavior is more likely to occur in the
future when an SD is present but is less likely to occur when an S� is present. This

126 Chapter 7

is the definition of stimulus control. It is important to remember that the presence
of an SD does not cause a behavior to occur; it does not strengthen a behavior.
Rather, it increases the likelihood of (or evokes) the behavior in the present situa-
tion because it was associated with reinforcement of the behavior in the past.
Reinforcement is what causes the behavior to occur when the SD is present.

Discrimination Training in the Laboratory
In the experiment reported by Holland and Skinner (1961), a hungry pigeon
stands in a small experimental chamber. The wall in front of the pigeon features
a round disk (called a key) and two lights, green and red. A pigeon has a natural
tendency to peck at objects. When it pecks at the key, a small amount of food is
delivered to an opening in the chamber. The food reinforces the behavior of peck-
ing the key.

How did Holland and Skinner bring the pigeon’s key-pecking behavior under
the stimulus control of the red light?

They turned on the red light (SD), and then whenever the pigeon pecked
the key, they delivered food (reinforcement). Sometimes they turned on the
green light (S�), and when the pigeon pecked the key, they did not deliver food
(extinction). Because of the process of discrimination training, the pigeon is more
likely to peck the key when the light is red and less likely to peck the key when
the light is green. The red light signals that key-pecking will be reinforced; the
green light signals that key-pecking will not be reinforced.

In similar experiments, a rat learns to press a lever in an experimental cham-
ber when the lever-pressing response is reinforced by food. Through discrimina-
tion training, a rat learns to press the lever when a certain audible tone is
presented and not to press the lever when a different tone is presented (Skinner,
1938).

Antecedent

Outcome: Pigeon pecks the key only when the red light is on.

Red light (SD) Pigeon pecks the key. Food is delivered.

ConsequenceBehavior

Green light (S ) Pigeon pecks the key. No food is given.

Antecedent Behavior

Outcome: Rat presses the lever only when the high-pitched tone is present.

High-pitched tone (SD)

Low-pitched tone (S )

Rat presses lever.

Rat presses lever.

Food is delivered.

No food is given.

Consequence

Stimulus Control: Discrimination and Generalization 127

Similarly, the recess bell develops stimulus control over children’s behavior in
elementary school. As soon as the bell rings, the students get up and go outside for
recess. This behavior is reinforced by playing and having fun. If the students got up
before the bell, the behavior would not be reinforced (the teacher would not let
them go outside to play). The recess bell is an SD for leaving the classroom because
the only time that leaving the classroom is reinforced is after the bell rings.

For each example of stimulus control in Table 7-1, identify the SD and the SA.

Answers are listed in Table 7-2.

Developing Reading and Spelling with Discrimination Training
Reading is a behavior that is developed through the process of stimulus discrimi-
nation training. Our reading behavior is under the stimulus control of the letters
and words we see on the page. If we see the letters DOG, we say “dog.” If we
said “dog” after seeing any other combination of letters, our response would be
incorrect. We learn to make correct reading responses through discrimination
training, typically when we are children.

Note that, in this example, the adult’s response “Wrong!” is a conditioned
punisher.

As we learn to read, we are able to discriminate the sound of each letter in the
alphabet, and we learn to read thousands of words. In each case, a particular letter is
associated with one sound, and a particular string of letters is associated with one word.
When we see a letter and make the correct sound, or see a written word and say the cor-
rect word, our behavior is reinforced by praise from teachers or parents. Thus, the letter
or the written word develops stimulus control over our reading behavior.

Describe how our behavior of spelling is developed through stimulus discrimina-
tion training.

TABLE 7-2 Discriminative Stimuli (SDs) and S-Deltas (S�s) for the Examples in Table 7-1

Example Behavior SD S�

1 Saying “I love you” Wife Coworkers

2 Stopping Red light Green light

3 Telling off-color jokes Friends Parents, teachers

4 Picking up the phone Phone rings No ring

5 Using the drill Light is on Light is off

Antecedent Behavior

Outcome: When the letters DOG are present, the child says “dog,” but the child does not say “dog” when any other
combination of letters is presented.

DOG (SD)

Another word (S )

The child says “dog.”

The child says “dog.”

Praise from teacher or parent.

No praise or teacher says “Wrong!”

Consequence

128 Chapter 7

In the case of spelling, the spoken word is the SD, and our response involves
writing or saying the letters that spell the word. When we write or say the letters
correctly, our spelling behavior is reinforced.

As a result of discrimination training, stimulus control develops over our spell-
ing behavior. Each particular word we hear (and each object or event we experi-
ence) is associated with only one correct spelling that is reinforced. Incorrect
spelling is not reinforced or is punished; thus, it no longer occurs.

Stimulus Discrimination Training and Punishment
Stimulus discrimination training may also occur with punishment. If a behavior is
punished in the presence of one antecedent stimulus, the behavior will decrease
and stop occurring in the future when that stimulus is present. The behavior may
continue to occur when other antecedent stimuli are present. For example, sup-
pose that when your soup is boiling, you put a spoonful in your mouth to taste it.
You burn your mouth, and as a result, you are less likely to put a spoonful of boil-
ing soup in your mouth in the future. However, you might still put soup in your
mouth before it is boiling or after it cools off, without burning yourself.

The boiling soup is a SD; it signals that tasting the soup will be punished.
Stimulus control has developed when you no longer try to taste soup that is boil-
ing. Consider another example. When you talk and laugh loudly in a library, the
librarian will tell you to be quiet or ask you to leave. However, talking and laugh-
ing loudly is not punished in many other situations (e.g., at a party or a ball
game). Therefore, the behavior of talking and laughing loudly is less likely to
occur in the library but continues to occur in other situations in which the behav-
ior is not punished.

The library is a SD for punishment that signals that loud talking and laughing
will be punished. Your behavior is under stimulus control when you no longer
laugh and talk loudly in the library.

Antecedent Behavior

Outcome: You are more likely to spell TREE when the teacher says “tree” and not when you hear any other
spoken word.

The teacher says, “Spell tree” (SD).

The teacher says, “Spell fish” (S ) or any other word.

You spell TREE.

You spell TREE.

The teacher gives praise.

The teacher says “Wrong.”

Consequence

Antecedent Behavior

Outcome: You are less likely to taste soup in the future when it is boiling.

Soup is boiling.

Soup is not boiling.

You taste a spoonful.

You taste a spoonful.

Painful stimulus (burned mouth)

No painful stimulus

Consequence

Stimulus Control: Discrimination and Generalization 129

The Three-Term Contingency

According to Skinner (1969), stimulus discrimination training involves a three-term
contingency, in which the consequence (reinforcer or punisher) is contingent on
the occurrence of the behavior only in the presence of the specific antecedent stim-
ulus called the SD. As you can see, a three-term contingency involves a relationship
among an antecedent stimulus, a behavior, and the consequence of the behavior.
Behavior analysts often call this three-term contingency the ABCs (antecedents,
behavior, consequences) of a behavior (Arndorfer & Miltenberger, 1993; Bijou,
Peterson, & Ault, 1968). The notation used to describe a three-term contingency
involving reinforcement is as follows:

SD � R � SR

where SD ¼ discriminative stimulus, R ¼ response (an instance of the behavior),
and SR ¼ reinforcer (or reinforcing stimulus). The notation for a three-term con-
tingency involving punishment is as follows:

SD � R � SP

In this case, SP ¼ punisher (or punishing stimulus).
As you can see, an antecedent stimulus develops stimulus control over

a behavior because the behavior is reinforced or punished only in the presence of
that particular antecedent stimulus. The same holds true for extinction. When
a behavior is no longer reinforced in a particular situation (in the presence of
a particular antecedent stimulus), the behavior decreases in the future only in that
particular situation.

Stimulus Control Research

Research has established the principle of stimulus control and explored its applica-
tion to help people change their behavior. For example, Azrin and Powell (1968)
conducted a study to help heavy smokers reduce the number of cigarettes they
smoked per day. The researchers developed a cigarette case that automatically
locked for a period of time (say, an hour) after the smoker took out a cigarette. At
the end of that period, the cigarette case made a sound to signal that the case
would open for another cigarette. The sound (auditory signal) was an SD that
signaled that trying to get a cigarette out of the case would be reinforced. Eventually,
stimulus control developed because the only time the smoker could get a cigarette

Antecedent Behavior

Outcome: You are less likely to laugh and talk loudly when you are in the library.

In a library

At a party

You laugh and talk loudly.

You laugh and talk loudly.

You are reprimanded.

You are not reprimanded.

Consequence

130 Chapter 7

was when the auditory signal (SD) was present. When the signal was not present, try-
ing to get a cigarette would not be reinforced because the case was locked.

Schaefer (1970) demonstrated that head-banging could be developed and
brought under stimulus control in rhesus monkeys. Schaefer was interested in
head-banging because this form of self-injurious behavior sometimes is seen in
people with intellectual disabilities. Through a procedure called shaping (see
Chapter 9), Schaefer got the monkeys to engage in head-banging and reinforced
this behavior with food. Discrimination training occurred in the following way.
Standing in front of the cage, Schaefer sometimes made verbal statements (SD) to
the monkey and sometimes said nothing (S�). When Schaefer said, “Poor boy!
Don’t do that! You’ll hurt yourself!” and the monkey hit its head, he delivered a
food pellet. When he did not provide the verbal stimulus and the monkey hit its
head, no food was provided. As a result, stimulus control developed, and the
monkey hit its head only when Schaefer made the statements (when the SD was
present). The verbal statements Schaefer used were similar to those sometimes
made by staff to people with intellectual disabilities who engage in self-injurious
behavior. Therefore, the study with monkeys had implications for the stimulus
control of self-injurious behavior in humans. Other researchers have evaluated
the stimulus control of self-injurious behavior (Lalli, Mace, Livezey, & Kates,
1998; Pace, Iwata, Edwards, & McCosh, 1986), other behaviors of people with
intellectual disabilities (Conners et al., 2000; Dixon, 1981; Halle, 1989; Halle &
Holt, 1991; Kennedy, 1994; Oliver, Oxener, Hearn & Hall, 2001; Striefel, Bryan, &
Aikens, 1974), and academic behavior and behavior problems of children (e.g.,
Asmus et al., 1999; Birnie-Selwyn & Guerin, 1997; Geren, Stromer, & Mackay,
1997; McComas et al., 1996; Richman et al., 2001; Ringdahl & Sellers, 2000;
Stromer, Mackay, Howell, McVay, & Flusser, 1996; Stromer, Mackay, &
Remington, 1996; Tiger & Hanley, 2004; Van Camp et al., 2000). Stimulus control
research has also been conducted with a variety of other populations and target
behaviors (Cooper, Heron, & Heward, 1987; 2007; Sulzer-Azaroff & Mayer,
1991). Chapter 16 discusses the application of stimulus control to help people
change their behavior.

FOR FURTHER READING
Stimulus Control and Rules

Stimulus control develops when a particular behavior is reinforced in the presence of an SD and the
behavior is then more likely to occur in the presence of the SD. Typically, the behavior must be rein-
forced in the presence of the SD a number of times before stimulus control develops. Sometimes, stim-
ulus control can develop more rapidly when rules are provided. A rule is a verbal statement specifying
the contingency, that is, telling the participant when (under what circumstances) the behavior will be
reinforced. Tiger and Hanley (2004) investigated the influence of rules on preschoolers’ behavior of ask-
ing for attention. In this study, preschoolers could only get their teacher’s attention when the teacher
wore a colored lei around his neck; they could not get the teacher’s attention when he was not wearing
the lei. The lei was the SD, asking for attention was the behavior, and getting attention was the rein-
forcer. Tiger and Hanley showed that when the preschoolers were given a rule (“When I am wearing
the red lei … I can answer your question …”), a greater degree of stimulus control developed than
when the rule was not provided. That is, when given the rule, the students were more likely to ask for
attention only when the teacher wore the lei.

Stimulus Control: Discrimination and Generalization 131

Generalization

In some cases, the antecedent conditions in which a behavior is strengthened
(through reinforcement) or weakened (through extinction or punishment) are
fairly specific; in others, the antecedent conditions are more broad or varied.
When the stimulus control of a behavior is more broad—that is, when the behav-
ior occurs in a range of antecedent situations—we say that stimulus generalization
has occurred.

Generalization takes place when a behavior occurs in the presence of stimuli
that are similar in some ways to the SD that was present during stimulus discrimi-
nation training (Stokes & Osnes, 1989). According to Skinner (1953a, p. 134),
“Generalization is … a term which describes the fact that the control acquired
by a stimulus is shared by other stimuli with common properties.” The more simi-
lar another stimulus is to the SD, the more likely it is that the behavior will occur
in the presence of that stimulus. As stimuli are less and less similar to the SD, the
behavior is less and less likely to occur in the presence of these stimuli. This is
called a generalization gradient (Skinner, 1957). Figure 7-1 presents an example
of a generalization gradient from a study by Guttman and Kalish (1956). Guttman
and Kalish reinforced key-pecking in pigeons when the key was illuminated with
a certain wavelength of light. As a result, the light was an SD that developed

Re
sp

on
se

s

40

80

120

160

200

240

280

320

Wavelength (millimicrons)
480 500 520 540 560 580 600 620

• •
•

•
•

•

• •

•

•

•
• •

•

•

•

•Group I Group II

FIGURE 7-1 This graph shows two stimulus generalization gradients in which pigeons’ key-pecking was rein-
forced when a 550-millimicron light was illuminated (discriminative stimulus [SD]). Subsequently,
they pecked the key when similar wavelengths of light were presented. The more similar the light
to the original SD, the more likely the pigeons were to peck the key. (From Guttman, N., & Kalish,
H. I. [1956]. Discriminability and stimulus generalization. Journal of Experimental Psychology, 51,
79–88. Reprinted by permission of the author’s heir.)

132 Chapter 7

stimulus control over the behavior, and the pigeon pecked the key whenever the
light was on. The graph shows that the pigeon also pecked the key when similar
wavelengths of light were presented. As the wavelength of light became less simi-
lar to the SD, less key-pecking occurred. The generalization gradient shows that
the behavior generalized to stimuli that are similar to the SD.

Another type of generalization gradient was demonstrated by Lalli and collea-
gues (1998). They showed that the head-hitting of a 10-year-old girl with intellec-
tual disability was reinforced by adult attention. The presence of an adult was an
SD for the behavior. In this case, the generalization gradient was the distance of
the adult from the child. When the adult was right next to the child, she was
more likely to engage in head hitting. The farther away the adult was from the
child, the less likely she was to engage in head-hitting. Figure 7-2 shows the gen-
eralization gradient from the study by Lalli and colleagues (1998). Other research
by Oliver and colleagues (2001) showed that close proximity to the therapist was
related to increased aggression exhibited by a girl with intellectual disability.

Examples of Generalization
A first grader, Erin is learning to read with the use of flash cards. When she sees
the card with MEN on it, she says “men” and gets praised. The MEN flash card
is an SD for saying “men.” At the mall with her parents one day, Erin sees the
MEN sign on the door of the men’s bathroom and says “men.” Because the
MEN sign on the bathroom is similar to the MEN flash card that was the original

Pe
rc

en
ta

ge
o

f T
ot

al
R

es
po

ns
es

A
cr

os
s

Se
ss

io
ns

20

40

50

60

70

80

90

100

Distance (meters)

Stimulus Generalization Gradient
of Self-Injurious Behavior

<.5M 1.5 3.0 4.5 6.0 7.5 9.0

10

30

FIGURE 7-2 The percentage of total responses across sessions at a given distance during generalization tests.
The closer the child was to the adult who reinforced the problem behavior, the more likely the
child was to engage in the behavior. (From Lalli, Mace, Livezey, & Kates [1998], copyright © 1998
Society for the Experimental Analysis of Behavior. Reprinted by permission of the Society for the
Experimental Analysis of Behavior.)

Stimulus Control: Discrimination and Generalization 133

SD, we say that generalization has occurred; the response occurred in the pres-
ence of a different stimulus that shared the same properties as the original SD.
Now if Erin reads the word men anywhere that she sees the letters MEN (e.g., in
a book, on a door, in block letters, or in handwritten letters), we can say that gen-
eralization has occurred to all relevant stimuli. Stimulus generalization in this
case is a desirable outcome of training. Erin has learned to discriminate all the
different ways that the word MEN may be written.

Stimulus generalization has also occurred when a response occurs in different
circumstances—in a different context, at a different time, or with different people—
from those in which it was originally learned. For example, parents may teach
their young child to follow their instructions or comply with their requests. When
the parents make a request (SD), the child complies with the request (R), and the
parents praise the child (SR). When the child complies with novel requests the
parents make, stimulus generalization has occurred. The specific request may
be new, but it shares the relevant features of the SD present during discrimination
training: It is a request or instruction made by the parent. Requests made by the
parent are part of a stimulus class: antecedent stimuli that share similar features
and have the same functional effect on a particular behavior. Stimulus generaliza-
tion has also occurred when the child complies with the request or instruction of
another adult (e.g., a teacher), in another context, or at another time. If the child
complies with the requests of other adults, the stimulus class that has acquired
stimulus control over the child’s compliance comprises requests made by adults
(as opposed to just requests by parents).

As you see, a stimulus control can be quite specific, or it can be broader. If a
behavior is reinforced in the presence of only one specific antecedent stimulus,
stimulus control is specific; the behavior is more likely to occur only when that
stimulus is present in the future. If a behavior is reinforced in the presence of a
number of antecedent stimuli that share the same features (that are in the same
stimulus class), stimulus control is broader and the behavior is more likely to
occur when any one of the antecedent stimuli from that stimulus class is present
in the future. Generalization is associated with broad stimulus control, or stimulus
control by novel or untrained antecedent stimuli.

Consider the example of 4-year-old Millie, a girl with severe intellectual
disability who exhibits a self-injurious behavior. Specifically, when her mother is
in the room, she gets down on her hands and knees and bangs her head on the
floor. When Millie bangs her head, her mother goes to her and stops her from
engaging in the behavior by holding her and talking to her (i.e., by paying atten-
tion to her).

Describe the three-term contingency (the ABCs) involved in Millie’s head-banging.

The antecedent stimulus or SD is the presence of her mother. The behavior is
banging her head on the floor, and the reinforcing consequence is her mom’s atten-
tion (holding her and talking to her). Head-banging is under stimulus control of her
mom’s presence. When her sisters are in the room but her mom is not present, Millie
does not bang her head because the behavior is never reinforced by her sisters.

When Millie went to the hospital recently, she banged her head when she
was with the nurse. This is an example of generalization. The presence of the

134 Chapter 7

nurse is a novel antecedent stimulus, but is similar to the SD (her mom, an adult).
When Millie banged her head with the nurse, the nurse held her and talked to
her, just as her mother does. In this way, the nurse reinforced her behavior.
While in the hospital, Millie banged her head when other adults entered her
room; these adults also reinforced the behavior. However, when Millie was in the
hospital playroom with another child, but no adult was present, Millie did not
bang her head.

Why doesn’t Millie bang her head when the only person in the room is another
child?

Millie does not bang her head when only a child is present because the other
children do not reinforce the behavior; they ignore Millie when she bangs her
head. Therefore, a child is an S� for the behavior. The behavior is under the stimu-
lus control of the presence of an adult because only adults reinforce the behavior.

Some examples of stimulus generalization are provided in Table 7-3.

Antecedent Behavior

Outcome: Millie bangs her head only when an adult is present.

Adult in the room

Another child in the room (but no adult)

Head-banging

Head-banging

Attention

No attention

Consequence

TABLE 7-3 Examples for Self-Assessment (Stimulus Generalization)

1. Amy is learning to identify the color red. When her teacher shows her a red block, she can say “red.”
Generalization has occurred when she also says “red” when the teacher shows her a red ball, a red
book, or any other red object.

2. Scott stopped putting his feet on the good coffee table after his wife yelled at him for doing it. Gen-
eralization has occurred when he stops putting his feet on the coffee table even when his wife is not
home.

3. Sharon’s dog Bud did not beg for food from her because she never gave Bud food when he begged.
However, when Sharon visited relatives for the holidays, her relatives reinforced begging behavior by
giving Bud food. After the holidays, when they were back home, Bud also begged for food from
Sharon and her friends. Generalization had occurred.

4. Sharon trained her dog Bud not to go into the streets around her house by using punishment. She
walked Bud on a leash near the street; each time Bud stepped into the street, Sharon snapped the
dog collar. Eventually, Bud no longer stepped into the streets even when not on a leash; generaliza-
tion had occurred. The dog also did not walk into the streets around other people’s houses; this was
another instance of generalization.

5. You learn to drive your brother’s car (which has a manual transmission) with your brother present.
The behavior then generalizes to most other cars with a manual transmission.

Stimulus Control: Discrimination and Generalization 135

In each example in Table 7-3, identify the three-term contingency used to
develop stimulus control initially, and identify the stimulus class that controls
the behavior after generalization has occurred.

The answers are given in Appendix A.

Look at the comic in Figure 7-3. Describe how this comic provides an example
of generalization.

FIGURE 7-3 In this comic strip you see an example of generalization. The SD was the paper in Dagwood’s front
yard, but the behavior (getting the paper) generalized to papers in the neighbors’ front yards.
(Reprinted with Special Permission of King Features Syndicate.)

136 Chapter 7

Initially, Dagwood taught Daisy to bring in the newspaper using the following
three-term contingency:

The newspaper in Dagwood’s front yard is the SD. Generalization occurred when
Daisy also brought in newspapers from the neighbors’ front yards. The stimulus class
controlling the response was a newspaper in the front yard of any house. Dagwood
wanted the stimulus class to be only the newspaper in the front yard of his house.

Describe how Dagwood would do discrimination training with Daisy to estab-
lish the correct stimulus control.

Dagwood should give Daisy a treat only when she brings in his paper and he
should give her no treat (and maybe a punisher) when she brings in a neighbor’s
paper. Behavior modification researchers and practitioners are quite interested in
stimulus generalization. When they use behavior modification procedures to help
people increase a behavioral deficit or decrease a behavioral excess, they want the
behavior change to generalize to all relevant stimulus situations. A number of
researchers have discussed strategies for promoting generalization of behavior
change (Edelstein, 1989; Kendall, 1989; Stokes & Baer, 1977; Stokes & Osnes,
1989). These strategies are reviewed in Chapter 19.

CHAPTER SUMMARY

1. An antecedent stimulus is a stimulus that precedes
the occurrence of the behavior. An operant behav-
ior is under stimulus control when it is more likely
to occur in the presence of a specific antecedent
stimulus or a member of a specific stimulus class.

2. Stimulus control develops through a process of
stimulus discrimination training, in which the
behavior is reinforced in the presence of one stim-
ulus (or stimulus class) but is not reinforced when

other stimuli are present. The antecedent stimu-
lus that is present when a behavior is reinforced is
called a discriminative stimulus (SD); an antecedent
stimulus that is present when the behavior is not rein-
forced is called an S-delta (S�). Stimulus discrimina-
tion training may occur with reinforcement,
punishment, or extinction; therefore, the occurrence
or nonoccurrence of a behavior may be under stimu-
lus control. However, it is not the SD that causes a

Antecedent Behavior

Outcome: In the future, Daisy brings in the newspaper when it is delivered to the front yard.

The newspaper is in the front yard. Daisy brings the newspaper to the house. Dagwood gives her a treat.

Consequence

Antecedent Behavior

Outcome: Daisy brings in Dagwood’s paper but does not bring in the neighbors’ papers.

Paper in front of Dagwood’s house (SD)

Paper in front of a neighbor’s house (S )

Daisy brings in paper.

Daisy brings in paper.

Daisy receives a treat.

No treat; Dagwood says, “No. Bad dog!”

Consequence

Stimulus Control: Discrimination and Generalization 137

behavior to occur or not occur. Reinforcement,
extinction, and punishment are the processes respon-
siblefor theoccurrenceornonoccurrenceof abehav-
ior in specific antecedent situations.

3. A three-term contingency involves a discriminative
stimulus (SD), a response that occurs in the pres-
ence of the SD, and a reinforcing consequence that

follows the response in the presence of the
SD (SD � R � SR ).

4. When stimulus control is broad or when a behavior
occurs in the presence of novel antecedent stimuli
that are similar to the initial SD, we say that generali-
zation has occurred. Stimulus control generalizes to a
class of stimuli sharing a particular feature or features.

KEY TERMS

antecedent, 123
discriminative stimulus (SD), 126
generalization, 132

S-delta (S�), 126
stimulus class, 134
stimulus control, 125

stimulus discrimination
training, 126

three-term contingency, 130

PRACTICE TEST

1. What is an antecedent stimulus? Provide an
example. (p. 123)

2. What does it mean when we say that the effects
of reinforcement are situation-specific? (p. 123)

3. What is stimulus control? (p. 125)
4. Provide an example of stimulus control.

(pp. 124–126)
5. What is an SD? What is an S�? (p. 126)
6. Describe stimulus discrimination training. What

is the outcome of stimulus discrimination train-
ing? (pp. 126–128)

7. Provide an example of stimulus discrimination
training with reinforcement and an example
with punishment. (pp. 127–129)

8. Does an SD cause a behavior to occur? Explain.
(p. 127)

9. What is a three-term contingency? Provide an
example. (p. 130)

10. A hungry rat presses a lever and gets food only
when a green light is on. What is the green
light? What will happen to the rat’s behavior of
pressing the lever in the future? (p. 127)

11. What is stimulus generalization? (p. 132)
12. Provide an example of stimulus generalization.

(pp. 132–135)
13. Whatisa stimulusclass?Provideanexample.(p.134)
14. Provideanexampleinwhichstimulusgeneralization

would be desirable. Provide an example in which
generalization would be undesirable. (pp. 133–135)

15. Describe how you would use stimulus discrimi-
nation training to make generalization more
likely or less likely to occur. (pp. 134–137)

APPENDIX A
The Three-Term Contingency and Outcome of Generalization in Each Example from Table 7-3

1. Antecedent�! Behavior�! Consequence
Red block Amy labels the color red. Amy receives praise from

the teacher.

Outcome:

Red block Amy labels the color red.

138 Chapter 7

After generalization:

Any red object Amy labels the color red.

2. Antecedent�! Behavior�! Consequence
Wife is present Scott puts feet on good coffee table. Scott gets yelled at.

Outcome:

Wife is present Scott does not put his feet on the
good coffee table.

After generalization:

Wife is not present Scott does not put his feet on the
good coffee table.

3. Antecedent�! Behavior�! Consequence
Around the relatives Bud begs for food. Relatives give Bud food.

Outcome:

Around the relatives Bud begs for food.

After generalization:

Around Sharon and her friends Bud begs for food.

4. Antecedent�! Behavior�! Consequence
With the leash on near Sharon’s
house

Bud steps into the street. Sharon snaps the dog collar.

Outcome:

With the leash on near Sharon’s
house

Bud does not step into the street.

After generalization:

With the leash off near Sharon’s
house

Bud does not step into the street.

With the leash off near other
people’s houses

Bud does not step into the street.

5. Antecedent�! Behavior�! Consequence
In your brother’s car (with manual
transmission) with your brother
present

You drive the car correctly. You receive praise.

Outcome:

In your brother’s car with your
brother present

You drive the car correctly.

After generalization:

In another car with manual trans-
mission, in your brother’s absence

You drive the car correctly.

Stimulus Control: Discrimination and Generalization 139

This page intentionally left blank

Chapter Eight

Respondent Conditioning

Chapters 4-7 describe principles of operant conditioning: reinforcement,extinction, punishment, and stimulus control. This chapter discusses a differ-
ent type of conditioning: respondent conditioning. Operant behaviors are
controlled by their consequences; operant conditioning involves the manipula-
tion of consequences. In contrast, respondent behaviors are controlled (elicited)
by antecedent stimuli, and respondent conditioning involves the manipulation
of antecedent stimuli. Consider the following examples.

Examples of Respondent
Conditioning

Carla worked in a factory that made children’s toys. She operated a
machine that molded plastic parts for the toys. Plastic pieces were fed
into the machine on a conveyor belt. As each piece entered the
machine, the machine made a clicking noise and then a metal punch
in the machine came down to stamp the plastic. When the machine
stamped the plastic, a short blast of air from one of the pneumatic hoses
hit Carla in the face. It was not dangerous, but the blast of air made her
blink each time the machine stamped a part. Carla found that she began
to blink as soon as the machine made the clicking sound, just before it

blew the air in her face. After a few days, the maintenance crew fixed the machine so
that the blast of air no longer came from the pneumatic hose. Carla noticed that she
continued to blink each time the machine clicked, but that the blinking went away
after a few days. Carla’s blinking is an example of a respondent behavior, elicited
by the antecedent stimulus of a blast of air in the face. Because the clicking sound
immediately preceded the blast of air each time, Carla’s blinking was conditioned to
occur at the clicking sound. This is an example of respondent conditioning.

Julio got out of his last class at 9:30 P.M. He took the 9:40 P.M. bus and got home
at 10:00 P.M. When he got off the bus, he had to walk through a tunnel under the
train tracks to get to his house. Because most of the lights in the tunnel were broken,
it was usually dark as he walked through it. Since the beginning of the semester, a
number of incidents in the tunnel had startled or scared him: A large rat ran right

â–  What is respondent conditioning?

â–  What are conditioned emotional
responses?

â–  How does extinction of respondent
behavior occur?

â–  What factors influence respondent
conditioning?

â–  How is respondent conditioning
different from operant conditioning?

141

in front of him; some teenagers made threatening remarks to him; and a homeless
person, who seemed to be sleeping, suddenly jumped up and started cursing at
Julio as he walked by. On each occasion, Julio noticed that his heart was racing, his
muscles were tensed, and he was breathing rapidly. These bodily responses contin-
ued until Julio came out of the tunnel. After these incidents, Julio noticed these
same bodily responses each time he walked toward the tunnel: His heart started rac-
ing, his muscles tensed, and his breathing was more rapid. These responses did not
diminish until he was out the other side. Once inside the tunnel, he usually walked
quickly or ran to get out more quickly. The increase in heart rate, muscle tension,
and rapid breathing is an example of respondent behavior. The threatening events
in the tunnel initially elicited bodily responses that we call fear responses or anxiety.
Because these events happened in the tunnel, proximity to the tunnel now elicits the
same bodily responses in Julio. Proximity to the tunnel is an antecedent stimulus that
elicits a conditioned response (CR) we call fear or anxiety.

Defining Respondent Conditioning

Certain types of stimuli typically elicit specific types of bodily responses. Infants
engage in sucking responses when an object such as a nipple touches their lips.
A person blinks when a puff of air is directed at the eye. The pupil of the eye
constricts on exposure to bright light. Salivation occurs when food is in the
mouth. A person gags or coughs when a foreign object is in the throat. These
and other responses (Table 8-1) are called unconditioned responses (URs).
These responses are elicited by antecedent stimuli even though no conditioning
or learning has occurred. A UR occurs in all healthy people when an uncondi-
tioned stimulus (US) is presented. We say that an unconditioned stimulus (US)
elicits an unconditioned response (UR). Humans have evolved to respond to USs
because the URs have survival value (Skinner, 1953a; Watson, 1924).

TABLE 8-1 Examples of Unconditioned Responses in Humans

Unconditioned Stimulus Unconditioned Response

Object touches infant’s lips Sucking reflex

Food in mouth Salivation

Foreign object in throat Gag reflex

Stimulation in the throat Coughing

Puff of air in the eye Eyeblink

Bright light in the eye Pupil constriction

Painful stimulation to the body Rapid withdrawal (of hand from a hot stove, for example) and
autonomic arousal (fight or flight response)

Sudden, intense stimulation (loud noise) Startle reflex (increased heart rate, respiration,
muscle tension)

Sexual stimulation (postpuberty) Erection or vaginal lubrication

Blow to the patella tendon Knee jerk

(From Pierce, W. D., & Epling, W. F. [1995]. Behavior Analysis and Learning, p. 65. Copyright © 1995 Prentice-
Hall, Inc. Reprinted by permission of the author.)

142 Chapter 8

Identify the ways in which each of the URs listed in Table 8-1 may have survival
value.

â–  The natural tendency to suck allows an infant to eat when a nipple is
placed in the mouth.

â–  Salivation contributes to chewing and digesting food.
â–  Gagging when a foreign object is in the throat can keep a person from

choking.
â–  Coughing clears the throat of foreign objects.
â–  The natural tendency to blink when air or other matter approaches the eyes can

prevent foreign objects from getting into the eyes and prevent loss of sight.
â–  Pupil constriction in response to bright light helps protect the eyes, and

thus prevent loss of sight.
â–  Rapid withdrawal from painful stimulation can help a person keep from

getting hurt (burned, cut, and so on).
â–  Autonomic nervous system arousal involves bodily systems that prepare a person

for action (the fight or flight response), and thus may enable the person to escape from
a dangerous situation or engage in protective behavior (Asterita, 1985). The bodily
responses involved in autonomic arousal are listed in Table 8-2.

â–  The startle response includes the components of autonomic arousal that
prepare the body for action in a possibly dangerous situation.

â–  The responses involved in sexual arousal do not have survival value for the
individual, but they facilitate sexual behavior, which is necessary for survival of the
human species.

â–  Although the knee jerk reflex may not have direct survival value itself, it is a
component of a larger group of reflexes involved in postural control and muscle
coordination that contribute to normal motor functioning.

A UR is a natural reflexive action of the body that occurs when a US is
present. URs are common to all people. Respondent conditioning occurs when a
previously neutral stimulus (NS) is paired with a US (the NS and the US are pre-
sented together). As a result of this pairing, the NS becomes a conditioned stimu-
lus (CS) and elicits a conditioned response (CR) similar to the UR. A UR or CR
is called a respondent behavior.

TABLE 8-2 Bodily Responses Involved in Autonomic Nervous System Arousal

Increased heart rate

Increased respiration

Increased muscle tension

Increased blood flow to major muscles

Decreased blood flow to the skin

Secretion of adrenalin into the bloodstream

Increased sweating

Dry mouth

Pupil dilation

Decreased gastrointestinal activity

Respondent Conditioning 143

Respondent conditioning is also called classical conditioning (Rachlin, 1976)
or Pavlovian conditioning (Chance, 1988). The Russian scientist Ivan Pavlov
(1927) was the first to demonstrate this phenomenon. In his experiments, Pavlov
showed that dogs salivated when meat powder was placed in their mouths. This
is a demonstration that a US will elicit a UR. Pavlov then presented a NS (the
sound of a metronome) just before he put the meat powder into the dog’s
mouth. He presented the sound of the metronome and the meat powder together
a number of times. After this, he presented the sound of the metronome by itself.
He found that the dog now salivated to the sound of the metronome without the
meat powder in its mouth. The sound of the metronome became a CS because it
was paired a number of times with the meat powder (US).

Just about any stimulus can become a CS if it is paired a number of times
with a US. Consider the case of Julio. Proximity to the tunnel became a CS
because it was paired with the US (startling events in the tunnel). As a result,
proximity to the tunnel elicited the CR of autonomic arousal (commonly called
fear or anxiety) that was previously elicited by the startling and frightening events.

Identify the US, UR, CS, and CR in the example of Carla in the toy factory.

The US is the blast of air in her face. It elicits the UR of blinking. Because
the clicking sound from the machine was paired with each blast of air, the
clicking sound became a CS. Now the clicking sound elicits the blinking, which
has become a CR. Note that blinking is a CR when elicited by the CS but was
initially a UR when elicited by the US.

US (meat powder)Process

US is paired with a neutral stimulus (metronome).

UR (salivation)

CS (metronome)Outcome

Note that the process involves pairing the US and neutral stimulus a number of times.
The outcome of the pairings is that the neutral stimulus becomes a CS and elicits a CR.

CR (salivation)

Respondent Conditioning

Blast of air (US)Process

The blast of air is paired with the clicking sound.

Blinking (UR)

Clicking sound (CS)Outcome Blinking (CR)

Respondent Conditioning

144 Chapter 8

Timing of the Neutral Stimulus and
Unconditioned Stimulus

The timing of the NS and US is important if respondent conditioning is to occur.
Ideally, the US should occur immediately after the onset of the NS (Pavlov, 1927).
In the case of Pavlov’s dogs, the metronome is sounded, and within about half a
second, the meat powder is placed in the dog’s mouth. This timing increases the
likelihood that the metronome will become conditioned as a CS. If Pavlov put
meat powder in the dog’s mouth and then sounded the metronome, it is unlikely
that conditioning would occur. The possible temporal relationships between the
NS and US are shown in Figure 8-1 (adapted from Pierce & Epling, 1995).

In trace conditioning, the NS precedes the US, but the NS ends before the
US is presented. In the eyeblink example, you present the clicking sound and,
after the clicking sound has stopped, you present the puff of air.

In delay conditioning, the NS is presented and then the US is presented before
the NS ends. Take the example of eyeblink conditioning. Delay conditioning occurs
if a clicking sound is presented and a puff of air is presented before the clicking sound
has terminated.

Delay Conditioning

NS

US

Trace Conditioning

NS

US

Simultaneous Conditioning

NS

US

Backward Conditioning

NS

US

Time

FIGURE 8-1 These time lines show the temporal relationship between the neutral stimulus (NS) and uncondi-
tioned stimulus (US) for four types of respondent conditioning. The raised portion of each time line
indicates when the stimulus (US or NS) is presented. Note that the stimulus labeled NS becomes a
conditioned stimulus only after pairing with the unconditioned stimulus. (From Pierce, W. D., &
Epling, W. F. [1995]. Behavior Analysis and Learning, p. 65. Copyright © 1995 Prentice-Hall, Inc.
Reprinted by permission of the author.)

Respondent Conditioning 145

In simultaneous conditioning, the NS and US are presented at the same
time. The clicking sound and the puff of air occur simultaneously.

In backward conditioning, the US is presented before the NS. In our
example, the puff of air is directed at the eye and then the clicking sound is pre-
sented. In those circumstances, it is unlikely that the clicking sound will elicit an
eyeblink response.

Of these types of respondent conditioning, trace and delay conditioning, in
which the NS is presented first, generally are most effective. Backward condition-
ing is least likely to be effective. Perhaps the only case in which respondent condi-
tioning can occur without close temporal proximity of the NS and US is taste
aversion. Consider the following example.

Murphy drank a glass of milk that had gone bad. Although the milk tasted
normal, Murphy experienced overwhelming nausea and vomiting 15 minutes
after he drank it. Since this episode, milk does not taste good to Murphy when
he tries to drink it. The tainted milk in Murphy’s stomach was a US, and the UR
was nausea and vomiting. Because the US was paired with the taste of milk, the
taste of milk became a CS that elicited a CR similar to the nausea he initially
experienced. Murphy might not actually get sick when drinking milk again,
but the milk does not taste good and may produce a milder version of the initial
nausea. This type of respondent conditioning is called taste aversion (Garcia,
Kimeldorf, & Koelling, 1955).

Higher-Order Conditioning

What you have learned so far is that an NS can become a CS when it is paired
with a US. The CS then elicits a CR. This is the basic process of respondent
conditioning. Higher-order conditioning occurs when an NS is paired with an
already-established CS and the NS becomes a CS. Consider the example of
Carla’s eyeblink response. Once the clicking sound was paired with the blast of
air a number of times, the clicking sound became a CS for Carla’s eyeblink
response. Now if another NS is paired with the clicking sound, it can become a
CS also. For example, if a light flashed each time the clicking sound was made,
the light would eventually become a CS and would elicit blinking even in the
absence of the clicking sound. Higher-order conditioning depends on how well
established the CS is when it is paired with the NS.

Blast of air in the face (US)Process

US is paired with clicking sound.

Eyeblink (UR)

Clicking sound (CS)Outcome Eyeblink (CR)

First-Order Conditioning

146 Chapter 8

Conditioned Emotional Responses

Some types of CRs produced through respondent conditioning are called condi-
tioned emotional responses (CERs). This term was first proposed by Watson and
Rayner (1920), who used respondent conditioning procedures to condition a fear
response in a young child, 1-year-old Albert. Little Albert was not initially afraid of
a white laboratory rat; he did not cry or try to get away. The rat was an NS. Watson
and Rayner presented the rat to Albert and immediately hit a metal bar with a ham-
mer behind Albert’s head (Figure 8-2). The loud, unexpected sound produced by
the hammer on the metal was a US that elicited a startle response (UR) in Albert.
The startle response involves autonomic arousal, the same type of responses
involved in fear or anxiety. After they paired the presence of the rat and the loud
noise seven times in two sessions 1 week apart, the rat became a CS. The sight of
the rat now elicited the CER we would call fear (e.g., crying, autonomic arousal).

Clicking sound (CS)Process

CS is paired with a flash of light.

Eyeblink (CR)

Flash of light (CS)Outcome Eyeblink (CR)

Higher-Order Conditioning

FIGURE 8-2 Watson hits the bar to make a loud noise as Albert touches the rat. After the startling noise and the
rat are paired a number of times, Albert shows a fear response when later presented with the rat.

Respondent Conditioning 147

Note that Watson and Rayner’s experiment with Albert probably also involved
operant conditioning. Initially, Albert reached for the white rat and the experi-
menters made the loud, startling noise. As a result of the pairing of the loud
noise and the white rat, the rat became a conditioned punisher. The behavior of
reaching for the rat was weakened through punishment, and the behavior of
crawling away from the rat was strengthened through negative reinforcement
(escape). Also note that this type of research, in which a fear response is induced
intentionally, would not currently be considered ethical.

The process of respondent conditioning can develop CSs for positive (desirable)
CERs or negative (undesirable) CERs (Watson, 1924). The fear developed in little
Albert by Watson and Rayner is an example of a negative CER; others include
anger, disgust, and prejudice. In the same way, positive CERs (e.g., happiness,
love) can be elicited by CSs. Initially, an emotional response is a UR elicited by a
US, such as a baby’s response to a mother’s physical contact. The mother strokes
the baby’s face and the baby smiles, coos, and makes other responses indicating
positive emotion. Eventually these CERs are conditioned to the sound of the
mother’s voice or the sight of her face. Another example would be when a young
man smells the perfume usually worn by his girlfriend and it elicits a positive
emotional response. Positive, affectionate interactions and physical contact with the
girlfriend would be the US eliciting the positive emotional response; the perfume is
the CS because it is paired with the US. Therefore, even if the girlfriend is not pres-
ent, the smell of the perfume can elicit the same feelings (positive CER) that the
young man experiences when he is with his girlfriend.

Identify positive and negative CERs occurring in your life and the CSs that
elicit these emotional responses.

Although the notion of CERs has intuitive appeal, there can be some difficulty
in operationalizing or measuring the emotional responses. Some emotional
responses are overt, and thus easily observable; these include crying, smiling, other
facial expressions, and postures indicative of autonomic arousal or calmness.
Likewise, the physiological responses involved in autonomic arousal (e.g., heart
rate, muscle tension, galvanic skin response), although covert, are measurable with
appropriate instruments. For example, muscle tension may be measured by electro-
myographic (EMG) recording, in which electrodes are placed on the subject’s
skin. The galvanic skin response records the changes in electrodermal activity that
accompany autonomic arousal because of increases in sweat gland activity. Auto-
nomic arousal may also be detected by recording the skin temperature at the tips
of the fingers. Because the blood flow is directed away from the surface of the skin
during autonomic arousal, the temperature of the hands and fingers decreases.

However, other reported emotional reactions are not observable or measurable;
these include feelings such as happiness or love. There is no doubt that people
experience positive and negative emotions that cannot be observed directly. The dif-
ficulty is that because they cannot be observed independently, it is not clear what
responses are involved in the emotions people report. Most likely, people’s reports
of emotional responses are a joint function of the actual CER, the situation in
which it occurs, their interpretation of events, and the ways in which they have
learned to label overt and covert events.

148 Chapter 8

Extinction of Conditioned Responses

Extinction of a CR, called respondent extinction, involves the repeated presenta-
tion of the CS without presenting the US. If the CS continues to occur in the
absence of the US, the CR eventually decreases in intensity and stops. If Pavlov
continued to present the sound of the metronome (CS) but never paired the
metronome with the delivery of meat powder (US), the dog would salivate less
and less to the sound of the metronome; finally, the dog would not salivate at all
when it heard the metronome.

In the case of little Albert, the white rat was a CS that elicited a fear response
(CR) because the rat had been paired with a loud, startling noise (US). In this
case, respondent extinction would occur if the white rat were presented to Albert
numerous times without the US. Eventually, the presence of the white rat would
no longer elicit a fear response.

Describe how respondent extinction occurred for Carla in the toy factory.

When the maintenance crew fixed the pneumatic hose, the blast of air no
longer occurred immediately after the clicking sound the machine made when it
stamped a plastic part. Because the CS (clicking sound) continued to be presented
in the absence of the US (blast of air), the CR (eye-blinking) eventually stopped
occurring when the CS occurred.

How would you use respondent extinction to help Julio eliminate his fear of
walking through the tunnel at night?

You would have to present the CS and prevent the occurrence of the US.
In other words, because proximity to the tunnel is the CS, he would have to walk
through the tunnel without any frightening or startling events (US) occurring.
If nothing bad ever happened in the tunnel again, the tunnel would no longer elicit
the autonomic arousal (fear response). This would not be easy to accomplish
because you cannot control who is in the tunnel or what happens there. One
solution would be to convince the city to replace the lights in the tunnel. If the
tunnel were brightly lit, startling events would be less likely to occur and threatening
people would be less likely to hang around in the tunnel.

Spontaneous Recovery
After a period of respondent extinction, in which the CS is repeatedly presented
in the absence of the US, the CS does not elicit the CR. However, if the CS
is presented at a later time, the CR might occur again. For example, Pavlov
presented the sound of the metronome repeatedly without putting meat powder
in the dog’s mouth. Eventually, the dog quit salivating to the sound of the metro-
nome. However, when Pavlov presented the metronome later, the dog again
salivated, although to a lesser extent than before extinction. When the CS elicits
the CR after extinction has taken place, spontaneous recovery has occurred. The
magnitude of the CR usually is smaller during spontaneous recovery, and the CR
should again disappear if the US is not presented with the CS during spontaneous
recovery.

Respondent Conditioning 149

Discrimination and Generalization
of Respondent Behavior

Discrimination in respondent conditioning is the situation in which the CR is eli-
cited by a single CS or a narrow range of CSs. Generalization has occurred when
a number of similar CSs or a broader range of CSs elicit the same CR. If a person
is afraid of a specific dog or a specific breed of dog, for example, discrimination
has occurred. If a person is afraid of any type of dog, generalization has occurred.

Consider how discrimination develops in respondent conditioning. When a
particular stimulus (S1) is paired with the US, but similar stimuli (S2, S3, S4,
etc.) are presented without the US, only S1 elicits a CR. This is discrimination
training. Consider the example of Madeline, who was attacked by a German
shepherd. Since the attack, every time she walks by the yard with the German
shepherd, the sight of the dog (CS) elicits autonomic arousal or a fear response
(CR). However, when she walks past other houses with different dogs, she does
not have the fear response. The sight of the German shepherd developed into a
CS because of its pairing with the attack (US). The sight of other dogs did not
develop into CSs because they were never associated with attacks. Now only the
sight of a German shepherd elicits the fear response (CR).

Now consider how generalization might develop. Generalization is the
tendency for the CR to occur in the presence of stimuli similar to the CS that
was initially paired with the US in respondent conditioning. If S1 is paired with
the US but similar stimuli (S2, S3, S4, etc.) are never presented in the absence
of the US, the CR is more likely to generalize to these other stimuli. If Madeline
was attacked by the German shepherd but she never had encounters with friendly
dogs, her fear response would be more likely to generalize to other dogs that
are similar in some way to German shepherds (dogs of similar size, similar color,
similar shape). In this case, there was no discrimination training because similar
stimuli (other dogs) were not presented in the absence of the US.

Generalization can be enhanced if a number of similar stimuli are paired
initially with the US during respondent conditioning. If Madeline was unfortunate
enough to be attacked by a German shepherd, a golden retriever, a schnauzer,
and a terrier, her fear probably would generalize to almost all dogs. Because a vari-
ety of similar CSs (different dogs) were all paired with the US (being attacked),
generalization would be enhanced.

Factors That Influence Respondent
Conditioning

The strength of respondent conditioning depends on a variety of factors
(Pavlov, 1927), including the following:

â–  The nature of the US and CS
â–  The temporal relationship between the CS and US
â–  Contingency between the CS and US

150 Chapter 8

â–  The number of pairings
â–  Previous exposure to the CS

The Nature of the Unconditioned Stimulus
and Conditioned Stimulus
The intensity of a stimulus influences the effectiveness of the stimulus as a CS or
a US. In general, a more intense stimulus is more effective as a US (Polenchar,
Romano, Steinmetz, & Patterson, 1984). For example, a stronger puff of air in
the eye is more effective than a weak puff of air as a US for an eyeblink response.
Likewise, a more painful stimulus is more effective than a less painful stimulus
as a US for autonomic arousal. A more intense stimulus also functions more
effectively as a CS; we say that the more intense stimulus is more salient.

The Temporal Relationship between the
Conditioned Stimulus and Unconditioned Stimulus
For conditioning to be most effective, the CS should precede the US. Therefore,
delay conditioning and trace conditioning are most effective. It is impossible to say
what time interval between the CS and the US is optimal; however, the interval
should be short (e.g., less than 1 second). The exception is taste aversion. The nausea
and vomiting (UR) elicited by the tainted food (US) may occur many minutes after
the occurrence of the CS (the taste of the food) in taste aversion conditioning.

Contingency between the Conditioned
Stimulus and Unconditioned Stimulus
Contingency between the CS and US means that the CS and US are presented
together on every trial. When this occurs, conditioning is much more likely than
if the US is not presented after the CS in some trials or if the US occurs in some
trials without the CS. When the machine clicks every time before it sends a blast
of air into Carla’s face, the click is much more likely to develop into a CS than if
the click were followed only occasionally (e.g., one of ten times) by the blast of air
in Carla’s face. Likewise, if the blast of air when the machine stamped a plastic
part was only occasionally preceded by a clicking sound, the clicking sound
would be unlikely to develop into a CS.

The Number of Pairings
Although one pairing between a NS and a US often is sufficient to establish the
NS as a CS, more pairings of the CS and US produce stronger conditioning in
general. Consider a student in an experiment who receives a brief electric shock
to the arm (US) after a buzzer sounds (CS); the shock is painful but, as in any
behavioral experiment, not strong enough to harm the student. After one pairing,
the buzzer probably will elicit autonomic arousal (CR). However, if the buzzer
and the shock are paired a number of times, the autonomic arousal will be stron-
ger and extinction will take longer to occur; that is, when the US is not presented,
the CS elicits the CR more times before the CR stops occurring. Even though
more pairings produce stronger conditioning, Rescorla and Wagner (1972)

Respondent Conditioning 151

demonstrated that the first pairing produces the strongest conditioning; the addi-
tional conditioning caused by each subsequent pairing steadily decreases. For
example, suppose that a big black crow screeches loudly as it flies by a young
child’s head. As a result, he experiences a fear response each time he sees a crow.
The first pairing of the crow (CS) and the attack (US) establishes the crow as a
CS that elicits the fear response (CR). If a crow swoops and screeches at the
child again, it may strengthen the child’s fear response, but the increase will not
be as great as the fear response produced by the first attack. Each additional attack
would increase the child’s fear by a progressively smaller amount.

Previous Exposure to the Conditioned Stimulus
A stimulus is less likely to become a CS when paired with a US if the person has
been exposed to that stimulus in the past without the US. For example, 2-year-old
Grace spends a lot of time around the family dog, Knute, and nothing bad ever
happens. As a result of this exposure to Knute, it is unlikely that Knute will become
a CS for a fear response from Grace if he accidentally knocks her down. However,
imagine that Grace’s friend Paula comes over and sees Knute for the first time.
If Knute accidentally knocks Paula down, it is more likely that Knute will become
a CS for a fear response because Paula had no previous exposure to Knute.
In the example of Knute and Paula, identify the US, CS, UR, and CR.

Getting knocked down by Knute is a US that elicits a UR of autonomic
arousal (fear response) in Paula. Knute is the CS because his presence was paired
with the US. As a result, Knute will elicit a fear response (CR) in Paula the next
time she sees him.

FOR FURTHER READING
Respondent Conditioning and Conditioned Punishers

Respondent conditioning is the process of pairing a neutral stimulus (NS) with an unconditioned stimulus.
Conditioned reinforcers and conditioned punishers are established through a respondent conditioning
process. An NS is paired with a reinforcer to produce a conditioned reinforcer or an NS is paired with a pun-
isher to produce a conditioned punisher. Research conducted in the 1960s demonstrated a number of factors
related to the development of conditioned punishers. For example, Evans (1962) showed that when a tone
was paired with shock, the tone functioned as a conditioned punisher for bar-pressing by laboratory rats.
Evans showed that, when the tone preceded the shock (trace conditioning), the tone was a more effective
punisher than when the tone followed the shock (backward conditioning). In another study, Hake and Azrin
(1965) showed that when a clicking sound was paired with shock, the clicking sound functioned as a condi-
tioned punisher for key-pecking by pigeons. Hake and Azrin further showed that when the clicking sound was
paired with a more intense shock, the clicking sound became a more effective conditioned punisher.

Distinguishing between Operant and
Respondent Conditioning

From the preceding discussion, it should be clear that respondent conditioning
and operant conditioning are distinct processes and that respondent and operant
behaviors include different types of responses (Michael, 1993a). A respondent

152 Chapter 8

behavior is a UR or CR elicited by an antecedent stimulus. Respondent behaviors
are bodily responses that have a biological basis. Operant behavior is controlled by
its consequences. Although it may be under the stimulus control of a discrimina-
tive stimulus (SD), an operant response is not elicited by an antecedent stimulus.
An operant response is emitted by the individual in specific antecedent situations
because it has been reinforced in the same or similar situations.

On Terms: the Difference Between Elicit and Evoke

We say that respondent behavior is elicited by an antecedent stimulus:

â–  a US elicits a UR as an unconditioned reflex
â–  a CS elicits a CR because the CS was paired with a US

We say that operant behavior is evoked by an antecedent stimulus or event:

â–  an SD evokes a behavior because the behavior had been reinforced in its presence
â–  an EO evokes a behavior because it increases the value of the reinforcer produced by the behavior

Respondent conditioning occurs when an NS acquires the power to elicit a
CR because the NS has been paired with a US. Respondent conditioning simply
involves pairing two stimuli: the NS and US. The outcome of respondent condi-
tioning is the development of a CS from a previously neutral stimulus. Operant
conditioning occurs when a specific response in a particular stimulus situation is
followed reliably by a reinforcing consequence. That is, operant conditioning
involves a contingency between a response and a reinforcer in specific circum-
stances. The result of operant conditioning is that the behavior is more likely to
occur in the future in circumstances similar to those in which the behavior was
reinforced. To describe this, we say that the circumstances in which the behavior
was reinforced develop stimulus control over the behavior or evoke the behavior.

Respondent extinction occurs when the CS is no longer paired with the US.
As a result, the CS no longer elicits the CR. Extinction of an operant behavior
occurs when the behavior no longer results in a reinforcing consequence and, as
a result, the behavior stops occurring in the future.

Operant and respondent behaviors can occur together in the same situation.
When the big black crow swoops down at the young child in the backyard and
screeches loudly, both respondent and operant behaviors are likely to occur. The
attack by the crow elicits autonomic arousal, and the child screams and runs to
his father, who is sitting in the yard and reading the paper (Figure 8-3). Although
autonomic arousal is a respondent behavior elicited by the crow, screaming and
running to the father are operant behaviors that result in comforting and attention
(positive reinforcement) and escape from the crow (negative reinforcement).

Consider the example of Carla in the toy factory. The clicking sound from
the machine before the blast of air is a CS that elicits an eyeblink response (CR)
because the clicking sound was paired with the blast of air. This is respondent
conditioning. After a while, Carla learned to move her head to the side as soon
as she heard the clicking sound. By doing so, she avoided the blast of air in the
face. Moving her head to the side is an operant behavior that is reinforced by its
consequence (avoiding the blast of air). The clicking sound is an SD that develops

Respondent Conditioning 153

stimulus control over (evokes) the behavior of turning her head. The behavior is
reinforced only when the clicking sound occurs. At any other time there is no
blast of air, and turning her head would not be reinforced.

Once she has learned to turn her head every time she hears the clicking
sound, respondent extinction occurs. She still hears the clicking sound, but the

FIGURE 8-3 When the crow swoops at the child, two types of behavior occur. The fear response (autonomic
arousal) is a respondent behavior (an unconditioned reflex); running to his father is an operant
response (reinforced by its consequences – escape from the crow and father’s comforting).

US (crow screeches, swoops down at the child)Process

US is paired with the sight of the crow.

UR (autonomic arousal)

CS (the sight of the crow)Outcome CR (autonomic arousal)

Respondent Conditioning

Operant Conditioning

Antecedent Behavior

Outcome: Child is more likely to run to his father when he sees a crow in the backyard.

Crow swoops and screeches. Child runs to his father. Father provides comfort.

Child escapes from crow.

Consequence

154 Chapter 8

blast of air doesn’t hit her in the face anymore. As a result, she stops blinking
(CR) when the clicking sound occurs (CS).

Identify the operant behavior and the respondent behavior in the example of
Julio and the dark tunnel.

The respondent behavior is the autonomic arousal elicited by proximity to the
tunnel. Proximity to the tunnel became a CS because startling or frightening events
(US) occurred in the tunnel. The operant behavior is walking quickly or running
through the tunnel. This behavior is reinforced by more rapid escape from the tun-
nel; in other words, this is negative reinforcement. Once Julio is out of the tunnel,
the autonomic arousal subsides. Therefore, the behavior is also negatively reinforced
by termination of the aversive physiological state of autonomic arousal.

Respondent Conditioning and Behavior
Modification

Most behavior modification procedures are designed to change operant behaviors
because operant behaviors make up the majority of behaviors that people target
for change. However, some types of respondent behaviors are also troublesome to
people and thus targeted for change. Most often, the types of respondent behaviors
that people want to change are CERs that interfere with normal functioning.

Respondent Behavior

Operant Behavior

SD (clicking sound) R (turn head) SR (avoid air in face)

CS (clicking sound) CR (eyeblink)

Respondent Behavior

Operant Behavior

Tunnel opening Running through the tunnel Escape from tunnel and escape from
autonomic arousal

SD R SR

CS (sight of tunnel) CR (autonomic arousal, fear response)

Respondent Conditioning 155

Thus, some people experience significant discomfort as a result of anxiety (e.g., anx-
iety about public speaking or anxiety in sexual situations). Sometimes, the auto-
nomic arousal elicited by the feared stimulus is so severe that the person alters his
or her life to avoid it; for example, a person with fear of heights may refuse to drive
over a particular bridge. Chapter 24 describes behavior modification procedures to
help people alter respondent behaviors involving fear and anxiety.

CHAPTER SUMMARY

1. In respondent conditioning, a previously neutral
stimulus (NS) becomes a conditioned stimulus
(CS) when it is paired with an unconditioned
stimulus (US). The CS elicits a conditioned
response (CR) similar to the unconditioned
response (UR) elicited by the US. Respondent
conditioning is most effective when the NS
immediately precedes the US. Higher-order con-
ditioning can occur when a NS is paired with an
already-established CS. Respondent behaviors
involve bodily responses that have survival value.

2. One type of respondent behavior is a conditioned
emotional response (CER). CERs may be nega-
tive (such as fear and anxiety) or positive (such as
happiness).

3. Respondent extinction occurs when the CS is
presented numerous times in the absence of the
US. As a result, the CS no longer elicits a CR.

4. Factors that influence respondent conditioning
include the intensity of the US or the CS, the
temporal relationship between the CS and the US,
the contingency between the CS and the US,
the number of pairings, and the person’s previous
exposure to the CS.

5. Respondent conditioning occurs when a NS is
paired with a US and the NS becomes a CS
that can elicit a CR. Operant conditioning occurs
when a behavior is reinforced in the presence of
an SD and the behavior is then more likely to
occur in the future when the SD is present.

KEY TERMS

backward conditioning, 146
conditioned emotional response

(CER), 147
conditioned response (CR), 143
conditioned stimulus (CS), 143
delay conditioning, 145
higher-order conditioning, 146

operant behavior, 141
operant conditioning, 141
respondent behavior, 141
respondent conditioning, 141
respondent extinction, 149
salient, 151
simultaneous conditioning, 146

spontaneous recovery
(respondent), 149

trace conditioning, 145
unconditioned response (UR), 142
unconditioned stimulus

(US), 142

PRACTICE TEST

1. Identify the terms signified by the following abbre-
viations: US, UR, CS, and CR. (pp. 142–143)

2. What is an unconditioned stimulus? Provide
examples. (pp. 142–143)

3. What is an unconditioned response? Provide
examples. (pp. 142–143)

4. Describe how a neutral stimulus (NS) becomes
a conditioned stimulus. What is this process
called? (p. 143)

5. What is the outcome of respondent conditioning?
(p. 143)

6. The timing of the NS and the US in respondent
conditioning is important. There are four possi-
ble temporal relationships between the NS and
US: delay conditioning, trace conditioning,
simultaneous conditioning, and backward con-
ditioning. Describe each type of conditioning.
(pp. 145–146)

156 Chapter 8

7. Identify the most effective and least effective of
the four types of conditioning listed in Question
6. (pp. 145–146)

8. Describe higher-order conditioning. Provide an
example. (pp. 146–147)

9. What is a CER? Provide examples of positive
and negative CERs. (pp. 147–148)

10. Describe respondent extinction and provide an
example. (p. 149)

11. What is spontaneous recovery of respondent
behavior? Provide an example. (p. 149)

12. How does taste aversion differ from other types
of respondent conditioning? (p. 146)

13. How is discrimination of respondent behavior
developed? Provide an example. (p. 150)

14. How is generalization of respondent behavior
developed? Provide an example. (p. 150)

15. Identify and describe the five factors that influ-
ence respondent conditioning. (pp. 150–152)

16. Describe how respondent and operant behavior
may occur together in the case of a student’s fear
of public speaking. (pp. 152–155)

17. How would you use respondent extinction to
help a child overcome a fear of dogs? How
would you use positive reinforcement in this
same case?

Respondent Conditioning 157

This page intentionally left blank

Chapter Nine

Shaping

As detailed in Chapter 4, reinforcement is a procedure for increasing thefrequency of a desirable behavior. To use reinforcement, the desirable
behavior must already be occurring at least occasionally. If the person does not
exhibit a particular target behavior at all, you need other strategies to generate the
behavior. Shaping is one such strategy.

An Example of Shaping: Teaching
a Child to Talk

Shaping happens naturally with children everywhere. A young child
who has not yet learned to talk will engage in babbling; that is, the
child makes word sounds that mimic the parents’ language. Initially,
the parents get excited and pay attention to the babbling child. Attentive
parents smile, talk to the child, imitate the word sounds, and stroke
the child; this attention reinforces the babbling behavior. As a result,
the child babbles more and more. Eventually, the child makes sounds
such as “da,” “ma,” or “ba” that resemble familiar words (“dada,”
“mama,” or “ball”). Again the parents get excited and pay attention to
these recognizable word sounds. As a result, the child starts to make
these sounds more often. At the same time, the parents do not respond
as much to the simple babbling once the child begins to make familiar
word sounds. As this process continues, the child eventually puts the
sounds together to make words, such as “dada” or “mama,” and the
parents get excited and provide more attention while paying much less

attention to the fragmentary word sounds that the child made before. As a result, the
child says words more often and makes word sounds (or babbles) less often. Through-
out the process of developing language, over the course of many months, the parents
reinforce closer and closer approximations to real words. The shaping process starts
when the parents reinforce babbling. The random word sounds in babbling are
approximations to actual words. Each time the child makes a sound that is a closer
approximation to a word, the child gets more attention (reinforcement) from the
parent, and the child gets less attention for the previous approximations.

â–  How do you use shaping to get a novel
behavior to occur?

â–  What are successive approximations
to a target behavior?

â–  How are the principles of reinforcement
and extinction involved in shaping?

â–  How might shaping be used
accidentally to develop a problem
behavior?

â–  What steps are involved in the
successful use of shaping?

159

It is also important to recognize that parents not only shape their child’s
language; they also bring it under proper stimulus control. The parents reinforce
“ba” or “ball” when showing the child a ball. They reinforce “da” or “dada”
when the child is looking or pointing at his or her dad. Through the process of
shaping, the child learns to say words; through discrimination training, the child
learns to say the correct words, that is, words that are appropriate to the situation.

Defining Shaping

Shaping is used to develop a target behavior that a person does not currently
exhibit. Shaping is defined as the differential reinforcement of successive approx-
imations of a target behavior until the person exhibits the target behavior. Differ-
ential reinforcement involves the basic principles of reinforcement and
extinction. Differential reinforcement occurs when one particular behavior is rein-
forced and all other behaviors are not reinforced in a particular situation. As a
result, the behavior that is reinforced increases and the behaviors that are not rein-
forced decrease through extinction. (See Chapter 15 for a more detailed discus-
sion on differential reinforcement procedures.)

When shaping is used to develop language, the successive approximations or
shaping steps include babbling, word sounds, part words, whole words, strings of
words, and sentences. To begin shaping, you identify an existing behavior that is
an approximation of the target behavior. This is called the starting behavior, or
first approximation. You reinforce this behavior and, as a result, the person starts
to exhibit this behavior more often. You then stop reinforcing the behavior and,
as part of the subsequent extinction burst, novel behaviors typically begin to
appear. Now you start reinforcing a novel behavior that is a closer approximation
to the target behavior. As a result, the person starts to exhibit the new behavior
more often and exhibits the previous behavior less often. This process of differen-
tial reinforcement (reinforcement of a closer approximation and extinction of a
previous approximation) continues until the person finally exhibits the target
behavior.

Skinner (1938) used shaping to get laboratory rats to press the lever in an
experimental chamber, which was about 1 × 1 square foot. The lever looked like
a bar sticking out of one wall of the chamber. The rat could easily put a paw on
the lever and push it. The chamber also had a small opening in the wall where
food could be delivered. When the rat was first put in the chamber, it wandered
around and explored.

Describe how you would use shaping to get the rat to press the lever.

First, you choose the starting behavior or the first approximation. You might
decide to deliver a pellet of food each time the rat steps to the side of the
chamber where the lever is located. As a result, the rat spends most of its time on
this side of the chamber. Now you reinforce the next approximation and put the
previous approximation on extinction: You deliver a pellet of food only when the
rat is facing the lever. As a result, the rat faces the lever frequently. Now, when
the rat approaches or moves closer to the lever, you deliver the food pellet. Next,
you deliver a pellet of food only when the rat is close to the lever and rears up on

160 Chapter 9

its hind legs. Once the rat engages in this behavior consistently, you put it
on extinction and deliver a pellet of food only when the rat makes a movement
toward the lever. Once this behavior occurs frequently, you then go to the next
approximation and deliver a pellet of food only when the rat is touching the lever
with its paw. Because this behavior is reinforced, the rat touches the lever
frequently. Finally, you move to the last step and provide a pellet of food only
when the rat presses the lever. Now, whenever this hungry rat is put into the
experimental chamber, it will rear up and press the lever with its paw because
that is the behavior that has been reinforced. Shaping allows you to begin by
reinforcing a behavior that the rat engages in frequently (standing on one side of
the chamber) and ends up with getting the rat to engage in a behavior it has
never performed (pressing the lever).

Successive Approximations to Lever-Pressing

1. The rat moves to the side where the lever is located.
2. The rat faces the lever.
3. The rat approaches the lever.
4. The rat rears up on its hind legs.
5. The rat makes a movement toward the lever with a paw.
6. The rat touches the lever.
7. The rat presses the lever.

Although we have outlined seven shaping steps (successive approximations),
many more steps may be included when shaping the lever-pressing response in the
rat. For example, step 3, in which the rat approaches the lever, could be further
divided into two or three steps. The important point is that each step should be a
closer approximation to the target behavior than was the previous step.

Have you ever wondered how dolphins and other sea mammals at aquatic
parks learn to perform complex tricks? Their trainers use shaping to get the
animals to engage in these behaviors (Pryor, 1985). Using a fish to eat as an
unconditioned reinforcer and a clicking sound from a handheld clicker as a con-
ditioned reinforcer, the dolphin trainers can shape complex behaviors by starting
with natural behaviors that the dolphins engage in frequently. By reinforcing
successive approximations, they can get the dolphins to engage in behaviors they
have never previously exhibited (such as jumping out of the water and catching
rings on their noses).

How do the trainers establish the clicking sound as a conditioned reinforcer and
why do they need to use a conditioned reinforcer?

The trainers make the clicking sound each time they give the dolphin a fish
to eat as a reinforcer. Because the clicking sound is paired with this uncondi-
tioned reinforcer, it becomes a conditioned reinforcer. They use the conditioned
reinforcer because the trainer can make the clicking sound quickly and easily,
and the dolphin’s behavior can be reinforced immediately without the disruption
of stopping to eat the fish. When using shaping, timing is important. You want to
deliver the reinforcer at the exact instant that the correct approximation occurs;
otherwise, you might accidentally reinforce a different behavior. In addition, the

Shaping 161

conditioned reinforcer is used so that the dolphins don’t become satiated with
fish. Fed fish as a reinforcer, the dolphins would eventually become satiated,
and fish would no longer function as a reinforcer until the dolphins were hungry
again. For further discussion of shaping with animals, see Pryor (1985) and
Skinner (1938, 1951, 1958).

Applications of Shaping

O’Neill and Gardner (1983) described a couple of interesting examples of shaping
human behavior in a medical rehabilitation setting.

Getting Mrs. F to Walk Again
One case involved Mrs. F, a 75-year-old woman who had hip replacement
surgery. To walk independently again, she needed physical therapy (PT). Specif-
ically, she had to walk between two parallel bars while supporting herself with
her arms on the bars. However, Mrs. F refused to participate in the PT. Because
Mrs. F was not currently exhibiting the target behavior, O’Neill and Gardner
(1983) decided to use shaping. The target behavior was walking independently
with her walker. For a starting behavior, they wanted Mrs. F to go to the PT
room where the parallel bars were located. When Mrs. F arrived in the PT
room in her wheelchair, the therapist interacted warmly with her and gave her
a massage treatment (a pleasant experience for Mrs. F). As a result, going to the
PT room was reinforced, and Mrs. F now went there willingly each day. After a
few days, the therapist asked Mrs. F to stand up between the parallel bars for
1 second (a successive approximation to walking) before she could have her mas-
sage. Mrs. F stood up for 1 second and received her massage. The therapist
increased the duration to 15 seconds the next day, and Mrs. F stood at the paral-
lel bars for 15 seconds before receiving her massage (Figure 9-1). After Mrs. F
was successfully standing between the parallel bars, the therapist asked her to
take a few steps one day and then a few more another day until she was walking
the full length of the parallel bars. Eventually, Mrs. F was walking indepen-
dently with her walker and was discharged from the hospital. Because shaping
involves starting with a simple behavior that the person is already engaging in
and building up to the target behavior in small steps (successive approxima-
tions), the person can engage in a new target behavior or a target behavior that
she previously refused to do.

Getting Mrs. S to Increase the Time between Bathroom Visits
Another case that O’Neill and Gardner (1983) reported involved Mrs. S, a
32-year-old woman with multiple sclerosis. In the hospital, she often interrupted
her therapy program for bathroom visits because she had once been incontinent
(lost bladder control) in public and was worried that it might happen again.
She was going to the bathroom more than once per hour. In collaboration with
Mrs. S, O’Neill and Gardner decided to use shaping to help her increase the
time between bathroom visits. The target behavior was to wait 2 hours between
trips to the bathroom. They decided that the starting behavior would be to wait

162 Chapter 9

1 hour, because Mrs. S occasionally waited 1 hour between bathroom trips before
the shaping program was started. Mrs. S successfully met this goal for a few days
and received therapist approval and praise as a reinforcer. The next approximation
was to wait 70 minutes. After Mrs. S waited successfully for 70 minutes for a few
days, the duration was increased to 90, then 105, and finally, 120 minutes. It took
12 days and 5 shaping steps for Mrs. S to reach the target behavior of waiting
120 minutes between bathroom trips (Figure 9-2). When she was discharged from
the hospital, the average time between bathroom visits was 130 minutes. Months
after she left the hospital, Mrs. S reported that she was maintaining her treatment
gains and that her life had improved as a result.

As you can see from these examples, shaping can be used in the following ways:

1. Generating a novel behavior (language in a young child, lever-pressing in
the laboratory rat, tricks from the dolphin)

2. Reinstating a previously exhibited behavior (walking, which Mrs. F was
refusing to do)

3. Changing some dimension of an existing behavior (the time between
urination for Mrs. S)

In each case, the target behavior is novel in that the person is not currently
engaging in that particular behavior.

FIGURE 9-1 Mrs. F stands between the parallel bars as one of the successive approximations in the shaping
process to get to the target behavior of walking with her walker.

Shaping 163

Research on Shaping

Research shows that shaping has been used to generate a variety of target
behaviors in a variety of populations including high-level athletic performance
(e.g., Scott, Scott, & Goldwater, 1997), therapeutic exercise for headache control
(Fitterling, Martin, Gramling, Cole, & Milan, 1988), toileting in infants
(Smeets, Lancioni, Ball, & Oliva, 1985), compliance with medical interventions
by individuals with intellectual disabilities (Hagopian & Thompson, 1999; Slifer,
Koontz, & Cataldo, 2002), and children’s use of contact lenses (Mathews,
Hodson, Crist, & LaRoche, 1992).

Studies by Jackson and Wallace (1974) and Howie and Woods (1982) report
the use of shaping to modify a dimension of an existing behavior. Jackson and
Wallace worked with a 15-year-old girl who exhibited mild intellectual disability
and was socially withdrawn. She spoke at a voice volume (loudness) that was
barely audible. The target behavior was speaking with a normal voice volume.
Jackson and Wallace used a decibel meter to measure the loudness of her speech
and reinforced successive approximations (louder and louder speech) with tokens

•

•

• •

•

• •
•

• •

•• •
•

• •

•
• •

• •
•

•

•

• •
• • •

•

•
•

•

•
• •

Vo
lu

m
e

of
U

rin
e

(c
ub

ic
c

en
tim

et
er

s)
In

te
rv

al
b

et
w

ee
n

U
rin

at
io

n
(m

in
ut

es
)

Days

80

100

120

140

50

60

70

80

90

100

110

120

130

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18

Baseline Treatment

Goal

3

1

1

FIGURE 9-2 This graph shows the average number of minutes between urination (top) and the volume of urine
(bottom) for Mrs. S each day. The stair-step line indicates the goal (successive approximation)
that was established for Mrs. S each day. Notice that the time between urination increases during
shaping and is always above the goal line. Also notice that the volume of urine per urination
increased as the time between urination increased. The numbers above the data points indicate
the number of times Mrs. S was incontinent. (From O’Neill, G. W., & Gardner, R. [1983]. Behavioral
Principles In Medical Rehabilitation: A Practical Guide, p. 49. Springfield, IL: Charles C. Thomas.
Reprinted by permission of the author.)

164 Chapter 9

until the girl was speaking at a more normal voice volume. They attributed the
success of the shaping program partly to the use of the decibel meter, which
allowed them to detect, and thus reinforce, very slight increases (successive
approximations) in the loudness of speech (Figure 9-3). Other researchers used a
shaping procedure to increase the voice volume of two children with disabilities.
Figure 9-4 shows the multiple-baseline-across-subjects graph from this study
(Fleece et al., 1981); an increase in voice volume was observed for both subjects.

Howie and Woods (1982) used shaping to increase the frequency of spoken
words in adults receiving treatment for stuttering. As part of their treatment,
the subjects slowed down their rate of speech as they learned to speak without
stuttering. After the subjects’ speech was stutter-free, the authors used shaping to
increase the rate of speech (syllables per minute) back to a more normal level.
In their study, the shaping steps or successive approximations involved increases
of five syllables spoken per minute. Using shaping, all subjects increased their
rate of speech to normal levels in about 40–50 sessions.

Shaping of different topographies (new forms) of behavior has been reported
in a number of studies (Horner, 1971; Isaacs, Thomas, & Goldiamond, 1960;
Lovaas, Berberich, Perdoff, & Schaeffer, 1966; Wolf, Risley, & Mees, 1964).
In an early study, Wolf et al. (1964) used shaping to get a preschooler with disabil-
ities to wear his glasses. Before the shaping procedure was used, the child refused
to wear his glasses; if anyone tried to make him wear them, he threw the glasses

FIGURE 9-3 The psychologist uses a decibel meter in the process of shaping increased voice volume (louder
speech) as a child speaks. Each shaping step involves successively louder speech, as measured
by the decibel meter.

Shaping 165

on the ground. The researchers used food to reinforce successive approximations
to the target behavior of wearing his glasses. The successive approximations
included touching the glasses, picking up the glasses, putting the glasses up to his
face, and finally putting on the glasses. By the end of the study, the child was
wearing his glasses regularly.

Horner (1971) worked with Dennis, a 5-year-old child with intellectual
disability. Dennis had a condition called spina bifida, in which the spinal cord is

Vo
ic

e
Vo

lu
m

e

Sessions

2

0

4

6

8

10

12

20

•

•

•
•

•
•

•

•

• •
• ••

•
•

• •

•

•
•

•

•

Baseline Treatment Follow-Up
Franklin

5 10 15 20

Usually
Screaming

Normal

Usually
Inaudible

2

0

4

6

8

10

12

20

•

•

• • • ••
•

•
•

• • • • • •• • • •
•

•

Baseline Treatment Follow-Up
Crystal

Usually
Screaming

Normal

Usually
Inaudible

4 Month Follow-Up1 Month Follow-Up

FIGURE 9-4 This graph shows the increase in voice volume for two students once shaping treatment was
implemented. For both children, the voice volume increased to normal levels and stayed that way
1 and 4 months after shaping was used. This graph illustrates a multiple-baseline-across-subjects
research design. Notice that the treatment (shaping) was implemented at a different time for each
subject, and that each subject’s behavior changed only after the treatment was implemented.
(From Fleece, L, Gross, A., O‘Brien, T, Kistner, J., Rothblum, E., & Drabman, R. [1981]. Elevation
of voice volume in young developmentally delayed children via an operant shaping procedure. Jour-
nal of Applied Behavior Analysis, 14, 351–355. Copyright © Society for the Experimental Analysis
of Behavior. Reprinted by permission of the Society for the Experimental Analysis of Behavior.)

166 Chapter 9

damaged before birth; as a result, use of the legs is limited. Dennis could crawl
but had never walked. Horner conducted two shaping procedures with Dennis.
In the first procedure, the target behavior was for Dennis to take ten steps while
holding himself up between parallel bars with his arms. This shaping procedure
included six stages. The first approximation was for Dennis to hold on to the
parallel bars with both hands while sitting on a stool. Horner used drinks of root
beer as the reinforcer for Dennis as he successfully completed each stage in the
procedure. After Dennis could walk using the parallel bars for support, the second
shaping procedure was started. The target behavior in this shaping procedure was
for Dennis to take 12 steps using forearm crutches. The first approximation to the
target behavior was for Dennis to hold the crutches in the correct position; the
second was to stand up using the crutches, with support from the experimenter;
the third was to stand up with no support; and so on. After 10 shaping steps in
120 training sessions, Dennis reached the target behavior. After he learned to use
the crutches successfully, Dennis walked to and from all of his activities in the
state institution where he lived. As a result of the shaping procedures conducted
by Horner, Dennis learned a behavior (walking) that made him more indepen-
dent and improved the quality of his life. The successive approximations involved
in the two shaping procedures Horner (1971) used are listed in Table 9-1.

(continued)

TABLE 9-1 Successive Approximations in the Two Shaping Procedures Used by Horner (1971)

The steps in the successive approximation sequence to establish use of parallel bars were as follows.

Step 1. Sitting on stool and gripping left parallel bar with left hand and right parallel bar with right
hand.

Step 2. Step 1 plus pulling to a standing position on parallel bars and maintaining a standing
position long enough to drink 1 tablespoon of root beer.

Step 3. Step 1 and step 2 plus taking one step using parallel bars for support before getting the reinforcer.

Step 4. Same as step 3 except three steps must be taken using parallel bars for support before
getting the reinforcer.

Step 5. Same as step 3 except five steps must be taken using parallel bars for support before
getting the reinforcer

Step 6. Same as step 3 except ten steps must be taken using parallel bars for support before
getting the reinforcer

The steps included in the successive approximation sequence to establish use of crutches were as
follows.

Step 1. Crutches secured to hands by elastic bandages. Experimenter stands behind child. Reinforcer
is delivered for imitating the modeled response of placing the crutches on dots marked on floor
18 inches in front of and 18 inches from each side of center line bisecting starting point.

Step 2. Crutches secured to hands by elastic bandages. Experimenter stands behind child.
Reinforcer is delivered contingent on completion of step 1 and swinging his body to a
crutches-supported erect position, with total assistance provided by experimenter through
underarm pressure. Erect position maintained for 15 seconds before reinforcer delivery.

Shaping 167

FOR FURTHER READING
Shaping and Compliance with Medical Procedures

Medical procedures often require the patient to tolerate diagnostic tests or medical interventions that may
take some time to complete (e.g., magnetic resonance imaging [MRI] procedures). In addition, ongoing
medical regimens for disease control often require the patient to engage in specific behaviors, some-
times on a daily basis (e.g., glucose testing for diabetics). Research has shown that shaping can
promote the behavior required for successful completion of medical procedures or medical regimens.
For example, Slifer, Koontz, and Cataldo (2002) use shaping to help children acquire the behavior
necessary to undergo MRI procedures. In this study, the researchers used shaping to reinforce longer
and longer periods of lying motionless in an MRI machine. Hagopian and Thompson (1999) use shaping
to help a child with intellectual disability and autism participate in a treatment regimen for cystic fibrosis.
The child needed to inhale for 20 seconds through a mask connected to an inhaler to receive his medi-
cation but was refusing to do so. In the shaping procedure, the child received a reinforcer (praise, candy,
small toy) for inhaling through the mask for 5 seconds initially, and then for increasingly longer durations.

How to Use Shaping

As you can see from the preceding examples, many applications of shaping
are reported in the research literature. It is appropriate to use shaping when your
therapeutic goal is to develop a target behavior that the person is not currently

Step 3. Crutches secured to hands by elastic bandages. Experimenter stands behind child.
Reinforcer is delivered contingent on completion of step 1 and swinging his body to a
crutches-supported erect position, with assistance provided by experimenter through
pressure under the arms only to prime initial movements.

Step 4. Crutches no longer secured to hands by elastic bandages. Initial assistance is no
longer provided. Reinforcement is contingent on independently swinging his body to a
crutches-supported erect position.

Step 5. Reinforcement is contingent on completion of step 4, maintaining balance with
experimenter’s hand placed on child’s back, and placing crutches in forward position.

Step 6. Reinforcement is contingent on completion of step 5 plus swinging feet toward an imaginary
line connecting crutch tips, maintaining balance with experimenter’s hand on child’s back,
and placing crutches in forward position.

Step 7. Reinforcement is contingent on completion of step 6 plus one additional cycle of placing
crutches in forward position, maintaining balance with experimenter’s hand on child’s back,
and placing crutches in forward position.

Step 8. Reinforcement is contingent on completion of four cycles of placing crutches in forward
position, and so on, with gradual fading of experimenter support during balancing.

Step 9. Reinforcement is contingent on completion of eight cycles of placing crutches in forward
position, maintaining balance without experimenter assistance, and placing crutches in
forward position.

Step 10. Reinforcement is contingent on completion of 12 cycles of placing crutches, maintaining
balance, and so on, using crutches with forearm clamps (Lofstrand type) instead of crutches
providing underarm support.

TABLE 9-1 Successive Approximations in the Two Shaping Procedures Used by Horner (1971)
(Continued )

168 Chapter 9

exhibiting. Shaping is one of a number of procedures that can be used to achieve
this goal (see Chapters 10–12).

The following steps ensure appropriate use of shaping (see also Cooper,
Heron, & Heward, 1987, 2007; Martin & Pear, 1992; Sulzer-Azaroff & Mayer,
1991; Sundel & Sundel, 1993).

1. Define the target behavior. By defining the target behavior, you can
determine whether and when your shaping program is successful.

2. Determine whether shaping is the most appropriate procedure. If the person
already engages in the target behavior at least occasionally, you do not need to
use shaping; you can simply use differential reinforcement to increase the fre-
quency of the target behavior. Shaping is used for the acquisition of a novel
topography or a novel dimension of a behavior or to reinstate a behavior that
the person does not currently exhibit. However, other, more efficient behav-
ioral acquisition strategies (such as prompting, modeling, or instructions) may
be preferable. You do not need to use shaping if you can simply tell a person
how to engage in the target behavior, if you can show the person the correct
behavior, or if you can physically assist the person to engage in the correct
behavior. See Chapters 10–12 for a discussion of these other strategies.

3. Identify the starting behavior. The starting behavior or first approximation
must be a behavior that the person already engages in, at least occasionally.
In addition, the starting behavior must have some relevance to the target
behavior. In every example in this chapter, the starting behavior was
chosen because the behavior was already occurring and it was an approxi-
mation that could be built on to get to the target behavior.

4. Choose the shaping steps. In shaping, the person must master each step
before moving to the next step. Each step must be a closer approximation
to the target behavior than the previous step (successive approximation).
However, the change in behavior from one step to the next must not be
so large that the person’s progress toward the target behavior stalls. A mod-
erate behavior change from one step to the next is most appropriate. If the
shaping steps are too small, progress will be slow and laborious. There is
no easy rule for choosing the shaping steps. You must simply choose the
shaping steps with the reasonable expectation that once a particular step is
mastered, it will facilitate the behavior specified in the next step.

5. Choose the reinforcer to use in the shaping procedure. You must choose a
consequence that will be a reinforcer for the person participating in the
shaping procedure. The trainer must be able to deliver the reinforcer
immediately contingent on appropriate behavior. The amount of the rein-
forcer should be such that the person does not satiate easily. Conditioned
reinforcers (such as tokens or praise) often are useful to avoid satiation.

6. Differentially reinforce each successive approximation. Beginning with the
starting behavior, reinforce each instance of the behavior until the behavior
occurs reliably. Then start reinforcing the next approximation while no longer
reinforcing the previous approximation. Once this approximation occurs con-
sistently, stop reinforcing this behavior and begin reinforcing the next approxi-
mation. Continue with this process of differential reinforcement of successive
approximations until the target behavior is occurring and being reinforced.

Shaping 169

7. Move through the shaping steps at a proper pace. Keep in mind that each
approximation is a stepping stone to the next one. Once a person masters
one approximation (successfully engages in the behavior at least a few
times), it is time to move to the next approximation. Reinforcing one approx-
imation too many times may make it difficult to move to the next step; the
person may continue to engage in the previous approximation. At the same
time, if the person does not master one approximation, it may be impossible
or at least difficult to progress to the next step. Successful movement from
one step to the next may be facilitated by telling the person what is expected
or by cueing or prompting the appropriate behavior (O’Neill & Gardner,
1983; Sulzer-Azaroff & Mayer, 1991). For example, O’Neill and Gardner
told Mrs. F that she had to stand using the parallel bars for 1 second before
she would get her massage. They told her what they expected her to do to
earn reinforcement at any particular shaping step.

Shaping Guidelines

1. Define the target behavior.
2. Determine whether shaping is the most appropriate procedure.
3. Identify the starting behavior.
4. Choose the shaping steps.
5. Choose the reinforcer.
6. Differentially reinforce successive approximations.
7. Move through the shaping steps at a proper pace.

Shaping of Problem Behaviors

In certain circumstances, problem behaviors may be developed unintentionally
through shaping. In such cases, successive approximations of a behavior that is
not beneficial to the person are reinforced.

Consider the following example. Mrs. Smith was having trouble with her
4-year-old son Tommy, who was engaging in disruptive behavior. Mrs. Smith
runs a mail order business from her home. When she was busy, Tommy often
interrupted her and asked or demanded that she play with him. Because Tommy
was persistent, Mrs. Smith usually stopped what she was doing to play for a while.
The three-term contingency was as follows.

Mrs. Smith asked Tommy’s pediatrician what she could do about this.
He suggested that when Tommy demanded that she play with him, Mrs. Smith

Antecedent Behavior

Outcome: Tommy’s behavior of interrupting his mom when she is working is strengthened.

Mom is working. Tommy interrupts and demands that she play with him. Mom plays with him.

Consequence

170 Chapter 9

was to say, “I’ll play later, Tommy,” and to continue working; she was to ignore
Tommy’s further attempts to interrupt her.

What is the behavioral principle involved in this plan?

The pediatrician was suggesting that Mrs. Smith use extinction and learn not
to reinforce Tommy’s frequent demands on her. The first time that Mrs. Smith
used extinction, Tommy got upset. He ran into the other room and screamed
(extinction burst). Concerned for her son, Mrs. Smith followed him, calmed him
down, and then played for a few minutes. She tried extinction the next time
Tommy demanded that she play with him. Again, he screamed and ran into the
other room. Mrs. Smith followed him, calmed him down, and played with him
again, so that he would stop screaming.

What reinforced Mrs. Smith’s behavior of playing with Tommy when he screamed?

Mrs. Smith’s behavior of playing with Tommy when he screamed was
negatively reinforced because he stopped screaming.

Mrs. Smith began to notice that Tommy was screaming frequently to get her to
play with him. She decided to use the pediatrician’s advice and try to ignore this new
behavior. The next time Tommy screamed, Mrs. Smith stayed at her desk and
ignored the behavior. Tommy screamed for 3 minutes straight, and then Mrs.
Smith heard a crash. She ran into the other room and saw that Tommy had thrown
his monster truck toy against the wall (extinction burst). Tommy was still screaming
and sobbing. Mrs. Smith sat Tommy down and told him not to throw his toys and
that they would play later. She helped him pick up the pieces of the truck and put
it back together. She talked to Tommy until he settled down. See Figure 9.5 for
another example of shaping a problem behavior.

FIGURE 9-5 This comic shows an example of a problem behavior most likely developed through shaping. The
child has to repeat his requests for a snack multiple times before his mother gives in and buys
him the snack. It is likely that, over time, the child had to ask more and more times for snacks at
the store until his mother gave in. In this way, his mother reinforced a greater and greater fre-
quency of the behavior through shaping.

Shaping 171

Mrs. Smith went back to work and, a short time later, Tommy started to scream
again. When Mrs. Smith did not come into the room, he threw his toys again. Mrs.
Smith believed she could not ignore this behavior, so she ran into the room and scolded
Tommy. She made him sit on the couch while she lectured him on his inappropriate
behavior. By the time Mrs. Smith went back to the pediatrician, 2 weeks after the
previous visit, Tommy was screaming frequently and throwing his toys. His problem
behavior was much worse than before. Unknown to Mrs. Smith, she had used shaping
to develop a worse topography of the problem behavior.

Describe how shaping was used (inadvertently) to develop Tommy’s problem
behavior of screaming and throwing his toys.

Mrs. Smith unwittingly used differential reinforcement of successive approxi-
mations. Tommy’s starting behavior of interrupting and demanding was reinforced
by the attention from his mom when she played with him. Then she ignored his
interrupting and demanding behavior (extinction) and reinforced the behavior of
running into the other room and screaming. Next, she ignored running and
screaming (extinction) and reinforced the behavior of screaming and throwing
toys. Without knowing it, Mrs. Smith was reinforcing each new behavior problem
with her attention. Most likely, many severe problem behaviors that people
(especially children) exhibit are developed through a similar process of shaping.

Think of some examples of problem behaviors that may have been developed
through shaping.

One possible example is the behavior of showing off; in this case, the person has
to keep showing off more and more (doing more and more risky things) to continue
getting people’s attention (Martin & Pear, 1992). Another example is self-injurious
behavior, such as head-slapping, which may have started as a mild behavior and
grown more severe through shaping. Initially, when the child was upset and slapped
his or her head, the parents responded with concern (attention), which reinforced the
behavior. When the behavior continued, the parents tried to ignore it. However, the
child slapped harder, and the parents responded again with concern. This reinforced
the harder head-slapping. This process was repeated a few more times; thus, harder
and harder head-slapping was reinforced, until the behavior was causing injury.
Shaping may also play a role in arguments between spouses. Over the course of many
arguments, one spouse has to argue longer and harder and louder before the other
spouse finally gives in; thus, the more intense arguing is reinforced. Shaping probably
accounts for numerous types of problem behaviors; in each case, the people involved
have no idea that they are shaping these problem behaviors by their own actions.

The duration of a young child’s crying at night may be lengthened by
shaping. The child’s crying is reinforced by parents often as they come to the
room to calm her down. Eventually, the parents may try to ignore the crying, but
when it persists, they come into the room, thus reinforcing a longer duration of
the behavior. After trying but failing to ignore the crying a number of times, lon-
ger and longer durations of the behavior are reinforced, until eventually the child
may cry at night for an hour or more.

Much anecdotal evidence exists that shaping can develop problem behaviors
in people. However, no research documenting this conclusion is available because
it would be unethical to intentionally shape problem behaviors in people who did

172 Chapter 9

not previously exhibit them. Some research studies do show that shaping can be
used to create problem behaviors in laboratory animals.

For example, Schaefer (1970) used shaping with two rhesus monkeys to
develop head-banging behavior, in which the monkey raised its paw and slapped
itself on the head. Using food as the reinforcer, Schaefer shaped head-banging by
differentially reinforcing three successive approximations. In the first approxima-
tion, Schaefer delivered a food pellet whenever the monkey raised its paw. After
the monkey was consistently raising its paw, Schaefer put this behavior on extinc-
tion and started reinforcing the second approximation, raising its paw above its
head. After the monkey raised its paw above its head consistently, Schaefer no lon-
ger reinforced this behavior and reinforced only the target behavior, which was to
bring the paw down on top of the head. It took 12 minutes to shape head-banging
in one monkey and 20 minutes to shape head-banging in the other. The target
behavior looked quite similar to the self-injurious behavior sometimes exhibited
by people with developmental disabilities. This study documents that such
behaviors may occur as a result of shaping, at least in the rhesus monkey. It is
also possible that shaping may be responsible for the development of self-
injurious behavior in some people with developmental disabilities.

This study and others (Rasey & Iversen, 1993) demonstrate experimentally
that shaping can produce maladaptive behaviors in the laboratory. Clinical experi-
ence also suggests that shaping sometimes results in problem behavior in everyday
life. For example, one mother often screamed at her son to get him to obey her.
When she wanted him to do something around the house, she repeated the
request five to ten times and raised her voice until she was yelling at him.
It appeared that this behavior had developed through shaping.

Describe how the mother’s behavior of repeating requests and yelling at her son
was developed through shaping.

At first when she asked her son to do something, he obeyed right away. After
a while, he ignored the first request and obeyed her only after she repeated the
request. Before long, he ignored two or three requests and did what she asked
only after the fourth or fifth request. Eventually, he ignored repeated requests
and obeyed her only after she raised her voice and repeated the requests. Finally,
she was yelling at him and repeating requests many times before he obeyed her.
The son had shaped the behavior in his mother by differentially reinforcing her
behavior of repeating requests more and more loudly, until she was screaming.
It is important to recognize the power of shaping so that people can use shaping
correctly to develop beneficial target behaviors and can avoid the accidental
shaping of problem behaviors.

CHAPTER SUMMARY

1. Shaping is a behavioral procedure in which succes-
sive approximations of a target behavior are differ-
entially reinforced until the person engages in the
target behavior. Shaping is used to develop a target
behavior that the person does not currently exhibit.

2. Successive approximations (or shaping steps) are
behaviors that are increasingly more similar to the
target behavior.

3. Reinforcement and extinction are involved in
shaping when successive approximations to the

Shaping 173

target behavior are reinforced and previous
approximations are put on extinction.

4. Shaping may be used inadvertently to develop
problem behaviors. When a mild problem behav-
ior is put on extinction and the problem worsens
during an extinction burst, the parent may then
reinforce the worse behavior. If this process con-
tinues a number of times, the problem behavior
may become progressively worse through a pro-
cess of differential reinforcement of worse and
worse instances (more intense, more frequent, or
longer durations) of the behavior.

5. The following steps are involved in the successful
use of shaping.

a. Define the target behavior.
b. Determine whether shaping is the most

appropriate procedure.
c. Identify the starting behavior.
d. Choose the shaping steps (successive

approximations).
e. Choose the reinforcer to use in the shaping

procedure.
f. Differentially reinforce each successive

approximation.
g. Move at a proper pace through the shaping

steps.

KEY TERMS

differential reinforcement, 160 shaping, 160 successive approximation, 160

PRACTICE TEST

1. What is shaping? (p. 160)
2. When is it appropriate to use shaping? When

would you not use shaping? (p. 163)
3. Which two behavioral principles are involved in

shaping? Explain. (p. 160)
4. What are successive approximations? (p. 160)
5. Provide an example of the differential reinforce-

ment of successive approximations of a behavior.
(pp. 159–161)

6. Provide two examples (not from the chapter) of
shaping in everyday life. (pp. 159–164)

7. Provide an example (not from the chapter) of
how a problem behavior may be developed
through shaping. (pp. 170–173)

8. The starting behavior (or first approximation)
used in a shaping procedure has two basic char-
acteristics. What are they? (p. 169)

9. Why might it be useful to use conditioned rein-
forcers when conducting a shaping procedure?
(p. 169)

10. Describe how shaping and discrimination train-
ing are used in the development of language in
young children. (pp. 159–160)

11. Shaping may be used to establish a new topogra-
phy of a behavior or a new dimension of a behav-
ior. Explain this statement. Provide an example
of shaping a new dimension of a behavior.
(pp. 163–165)

12. Describe how an extinction burst may play
a role in shaping. Provide an example.
(pp. 160–171)

APPLICATIONS

1. Imagine you live in a house with a backyard. The
door to the backyard is in your family room. You let
your dog Felix out into the backyard a few times a
day. You decide you would like to teach Felix to
bump the doorknob on the back door with his nose

before you let him outside. Currently, whenever
Felix wants to go outside, he walks around the fam-
ily room and often walks past the back door.
Describe how you will use shaping to teach Felix
to bump the doorknob with his nose.

174 Chapter 9

a. What is your starting behavior?
b. What is your target behavior?
c. What will you use as a reinforcer during

shaping?
d. What are the successive approximations?
e. How will you use differential reinforcement

with each approximation?
f. What will you use as the natural reinforcer

for the target behavior once you reach it?
2. According to a much-told story, the students in

one of B. F. Skinner’s classes used shaping to get
Skinner to stand in the front corner of the class-
room when he lectured to the class. Let’s say
that you wanted to play a similar trick on one
of your professors. Assuming that the professor
moves around the front of the classroom at
least occasionally when he or she lectures and
assuming that student attention in class is a rein-
forcer for the professor, how would you use
shaping to get your professor to stand in one
corner of the classroom while lecturing?

3. One other application of shaping is a game that
can be educational and fun. Choose one person

who will be the trainer. Choose another person
whose behavior the trainer will shape. Call this
person the student. The trainer should have a
handheld clicker. The clicking sound will be
the reinforcer. The trainer and the student can-
not say anything during the shaping game. The
trainer decides on a target behavior but doesn’t
tell the student what it is. The game starts with
the student engaging in random behaviors. The
trainer tries to reinforce successive approxima-
tions to the target behavior until the student
exhibits the target behavior. The student must
respond according to the principle of reinforce-
ment; that is, the student begins to engage more
often in behaviors that are greeted by the sound
of the clicker. The trainer’s success depends on
how well he or she can choose and reinforce
successive approximations immediately as they
occur. This game is similar to the children’s
game of hot and cold, in which the child says
“hotter” as you move toward a target location
and “colder” as you move away.

MISAPPLICATIONS

1. Jody asked her father to teach her how to drive.
Her father had recently taken a class in behavior
modification and reasoned that because driving
was a novel behavior for Jody, she could be
taught how to drive by a shaping procedure.
What is the problem with this application of
shaping?

2. Every day, Mrs. Markle gives her second graders
a math worksheet to fill out. The worksheet con-
sists of five addition or subtraction problems. Mrs.
Markle noticed that Jake completes all five pro-
blems on his worksheet only once or twice a
week. She wanted Jake to complete all five pro-
blems every day. She decided that she would use
shaping with Jake to achieve this goal. What is
the problem with this application of shaping?

3. Dr. Williams, a school psychologist, was working
with an extremely socially withdrawn adolescent,

Jenny. Dr. Williams decided to use shaping to help
Jenny develop appropriate social skills. He identi-
fied the target behavior as making eye contact,
smiling, standing up straight, talking at a normal
voice volume, and nodding and paraphrasing
when the other person said something. Dr.
Williams was going to reinforce successive approx-
imations of this target behavior in therapy sessions,
in which he played the role of a classmate and
engaged in conversations with Jenny. In each ses-
sion, Dr. Williams and Jenny role-played four or
five short conversations. Before each role-play ses-
sion, Dr. Williams reminded Jenny which beha-
viors she should work on. As a reinforcer for
exhibiting the correct behavior in the role-play
sessions, Dr. Williams bought Jenny an ice cream
cone in the school cafeteria once a week. What is
the problem with this application of shaping?

Shaping 175

This page intentionally left blank

Chapter Ten

Prompting and Transfer
of Stimulus Control

You have already learned about shaping, a procedure for establishing desirablebehavior. This chapter discusses prompting and transfer of stimulus control,
which also are used to establish desirable behavior and to develop appropriate
stimulus control over desired behaviors (Billingsley & Romer, 1983).

An Example of Prompting and Fading:
Teaching Little Leaguers to Hit the Ball

Coach McCall was teaching first graders how to hit a baseball thrown
by a pitcher. Previously, the players had hit the baseball only off of a
tee. Luke was a good baseball player and a fast learner. Coach
McCall told Luke to stand in the batter’s box, to hold the bat back,
to start his swing a little before the ball got to the plate, to swing
level, and to watch the ball all the way to the bat. The assistant
coach, Dave, threw some pitches to Luke while coach McCall stood
nearby. Coach McCall praised Luke each time he hit the ball and
continued to give Luke instructions when he needed to improve his
performance. As Luke hit the balls successfully, the coach no longer
gave instructions but continued to praise him for each hit.

Next up was Tom. He listened to the same instructions that Luke heard but
could not hit the ball. To help him, coach McCall provided more assistance.
He pointed to where Tom should stand and gestured how the ball would come
in over the plate and where Tom should swing the bat. With this extra help,
Tom started to hit the ball and coach McCall praised him each time. Eventually,
Tom hit the ball without any extra help or instructions.

Matt watched and listened to coach McCall but still could not hit the ball.
To help Matt, coach McCall decided to show him exactly how to hit the ball.
Dave threw some pitches to coach McCall, who described the important aspects of
his own behavior as he hit them. After Matt listened to the instructions and watched
the coach hit the ball, he was able to hit the ball himself. Once Matt started to hit
the ball, coach McCall didn’t need to give him any further help (instructions or
modeling), but he still praised Matt each time Matt hit the ball correctly.

â–  What is prompting, and why is it used?

â–  What is fading, and why is it used?

â–  How do response prompts differ from
stimulus prompts?

â–  What are the different types
of response prompts?

â–  What does it mean to transfer stimulus
control, and how do you do it?

177

Finally, there was Trevor. Trevor watched and listened to everything coach
McCall was saying and doing, but he just couldn’t connect. Because Trevor
needed the most help, coach McCall stood behind him as he batted. He put his
hands over Trevor’s hands on the bat and helped Trevor swing the bat and con-
nect with the ball (Figure 10-1). After doing this a few times, coach McCall
backed off a little: He got Trevor positioned and started the swing with him, but
then let Trevor finish the swing himself. The coach then backed off a little more:
He got Trevor positioned and told him when to swing but let Trevor swing the bat
himself. After a few minutes, Trevor was hitting the ball independently, and all
that the coach had to do was provide praise each time.

Up to this point, Dave had been throwing easy pitches for the players to hit.
Dave was really close to the plate (he threw from behind a screen so he would
not get hurt when the kids hit the ball) and the pitches were slow and thrown
right over the plate. Once they could all hit the easy ones, Dave started to
throw pitches that were progressively more difficult to hit. First he threw them
from farther away. Then he threw them faster. And then he threw the pitches
in more difficult positions. He gradually increased the difficulty of the pitches
over the next four or five practices, and the players continued to hit the ball
successfully.

This example illustrates the behavior modification procedures called prompt-
ing and fading. All of the things that coach McCall did to help the players hit
the ball are prompts. With Luke, coach McCall provided a verbal prompt: He

FIGURE 10-1 The coach is using physical prompts, hand-over-hand guidance, to help Trevor hit the baseball.
Later he will fade the physical prompts and gradually eliminate the assistance until Trevor hits the
ball without any assistance.

178 Chapter 10

told Luke how to hit the ball correctly. With Tom, he gave a verbal and gestural
prompt: He gave instructions and motioned to Tom how to swing the bat. Coach
McCall provided a verbal prompt and modeling prompt for Matt: He told Matt
how to hit the ball and showed him the desirable behavior. Finally, for Trevor,
coach McCall provided a verbal and physical prompt. With the physical prompt,
he physically guided Trevor through the correct behavior until Trevor could do
it himself.

What Is Prompting?

As you can see, prompts are used to increase the likelihood that a person will
engage in the correct behavior at the correct time. They are used during discrimi-
nation training to help the person engage in the correct behavior in the presence
of the discriminative stimulus (SD). “Prompts are stimuli given before or during
the performance of a behavior: They help behavior occur so that the teacher can
provide reinforcement” (Cooper, Heron, & Heward, 1987, p. 312).

In this example, the SD is the ball approaching the batter. The correct
response is swinging the bat to connect with the ball, and the reinforcer is hitting
the ball and getting praise from the coach.

However, if the correct behavior is not occurring (if the player is not correctly
swinging the bat to hit the ball), the behavior cannot be reinforced. The function
of prompts is to produce an instance of the correct behavior so that it can be
reinforced. This is what teaching is all about: The teacher provides supplemental
stimuli (prompts) together with the SD so that the student will exhibit the correct
behavior. The teacher then reinforces the correct behavior so that it will eventually
occur whenever the SD is present (Skinner, 1968).

The use of prompts makes teaching or training more efficient. Coach McCall
could have simply waited for his players to hit the ball without any prompts and
praised them when they did so. But this trial-and-error process would have been
quite slow; some players might never have made a correct response. When coach
McCall used prompts, he increased the chances that his players would make a

Antecedent Behavior

Outcome: The batter is more likely to swing correctly and hit the ball thrown by the pitcher.

Pitcher throws the ball. Batter correctly swings the bat. Batter hits the ball and gets
praise from the coach.

Consequence

Antecedent Behavior

Pitcher throws the ball (SD).
Instructions are given (prompt).

Luke correctly swings the bat. Luke hits the ball and the
coach provides praise.

Consequence

Prompting and Transfer of Stimulus Control 179

correct response. For different players, he used different prompts (instructions,
gestures, modeling, and physical assistance) to get the correct response in the pres-
ence of the SD (the ball thrown by the pitcher).

What Is Fading?

Once the players were hitting the ball correctly, coach McCall faded his prompts.
Fading is the gradual elimination of the prompt as the behavior continues to
occur in the presense of the SD. Fading is one way to transfer stimulus control
from the prompts to the SD. Coach McCall gradually removed the prompts until
the batters hit the ball without any further assistance. That is, he stopped giving
instruction and he no longer had to model the behavior or provide physical
assistance to help the players hit the ball. Once the prompts were removed, the
behavior was under the stimulus control of the SD. When coach McCall was
using a physical prompt with Trevor, Trevor’s correct behavior was under the stim-
ulus control of the physical prompt. That is, he could hit the ball only because
the coach was helping him. But Trevor cannot have the coach physically assisting
him when he is batting in a game; he has to hit the ball on his own. Therefore,
teaching is not complete until prompts are completely faded (help is removed)
and the behavior is under the stimulus control of the natural SD.

Consider another example of prompting and fading. Natasha, a recent immi-
grant, is learning English in an adult education class. The class is learning to read
simple words. The teacher holds up a flash card with the letters CAR. When
Natasha does not respond, the teacher says “car,” and Natasha repeats the word
“car.” The teacher holds up the flash card again, and when Natasha says “car,”
the teacher says “Good!” The teacher then repeats this process with each of the
ten flash cards.

What Type of Prompt Is the Teacher Using?

When the teacher says the word on the flash card, this is a verbal prompt.
In this case, the verbal prompt is also a modeling prompt. The written word on
the flash card is the SD; saying the word (reading) is the correct response for

Antecedent Behavior

Outcome: Trevor hits the ball when it is pitched to him in the future.

Pitcher throws the ball (SD).
No more prompts.

Trevor swings the bat correctly. Trevor hits the ball and the
coach praises him.

Consequence

Antecedent Behavior

Flashcard is shown with the letters CAR (SD).
Teacher says “car” (prompt).

Natasha says “car.” Teacher praises Natasha.

Consequence

180 Chapter 10

Natasha. The verbal prompt helps Natasha make the correct response in the pres-
ence of the SD. But Natasha must be able to make the correct response when she
sees the written words without the prompt. To accomplish this, the teacher begins
to fade the verbal prompts. The second time through the set of flash cards, she
shows Natasha a flash card and, if she does not respond, the teacher says part
of the word as a prompt and Natasha says the whole word. The teacher shows
her the flash card again and Natasha then reads the word without the prompt.
The teacher provides praise for each correct response. The next time through the
flash cards, if Natasha cannot read a word, the teacher makes the sound of the first
letter in the word as a verbal prompt and Natasha says the whole word. The
teacher then shows her the flash card again, and she reads the word without a
prompt. Eventually, Natasha will read the words on the flash cards without any
prompts. At this point, her reading behavior is under the stimulus control of the
written words, not the verbal prompts (Figure 10-2).

Engaging in the correct behavior without prompts is the goal of prompting and
fading. Ultimately, the SD must have stimulus control over the behavior. Prompting
and fading help to establish appropriate stimulus control. Prompting gets the correct
behavior to occur; fading transfers stimulus control to the natural SD.

FIGURE 10-2 The teacher shows the students a flash card with a word on it (SD). If the students cannot make
the correct response (read the word), she provides a verbal prompt (says the word). Eventually,
she will fade the prompt, and the students will read the word presented on the flash card without
any assistance.

Antecedent Behavior

Outcome: Whenever Natasha sees the letters CAR, she says “car.”

CAR (SD)
No prompt

Natasha says “car.” Teacher praises Natasha.

Consequence

Prompting and Transfer of Stimulus Control 181

In this example, the teacher faded the prompts in three steps. First, she pre-
sented the flash card and said the whole word. The second time, she said the first
part of the word. The third time, she presented the flash card and pronounced the
first letter of the word. Finally, she presented the flash card and said nothing. Each
step was a gradual elimination of the prompt. By gradually eliminating the prompt,
the teacher transferred stimulus control from the prompt to the SD (written word).
In fading, transfer of stimulus control happens because the SD is always present
when the correct response is emitted and reinforced, whereas the prompt is
removed over time. As you can see, the prompting and fading facilitated stimulus
discrimination training: They made it possible for the correct reading response to
occur in the presence of the SD (word on the flash card) and be reinforced.

Types of Prompts

As we have seen, a prompt is an antecedent stimulus or event used to evoke the
appropriate behavior in a particular situation. Various types of prompts are used
in behavior modification; the two major categories are response prompts and
stimulus prompts (Alberto & Troutman, 1986; Cooper et al., 1987).

Response Prompts
A response prompt is the behavior of another person that evokes the desired
response in the presence of the SD. Verbal prompts, gestural prompts, modeling
prompts, and physical prompts are all response prompts.

Verbal Prompts When the verbal behavior of another person results in the correct
response in presence of the SD, this is a verbal prompt. It is a verbal prompt when
you say something that helps the person engage in the correct behavior. When
Natasha was learning to read, the teacher showed her the flash card with the word
CAR and said “car” (a verbal prompt). By saying “car,” she prompted Natasha to
make the correct response. When coach McCall told Luke how to swing the bat
to hit the ball, he was providing a verbal prompt (instruction). The verbal prompt
led to the desired behavior (swinging the bat correctly) in the presence of the SD

(the ball thrown by the pitcher). Any verbal statement from another person may act
as a verbal prompt if it makes the correct behavior more likely to occur at the right
time. Verbal prompts may include instructions, rules, hints, reminders, questions, or
any other verbal assistance. Verbal prompts can be effective because people have
a history of reinforcement for following instructions (rules, reminders, etc), and as a
result, develop a generalized instruction following repertoire. Therefore instructions
have stimulus control over instruction-following behavior.

Gestural Prompts Any physical movement or gesture of another person that
leads to the correct behavior in the presence of the SD is considered a gestural
prompt. However, if the person demonstrates or models the entire behavior, it is
considered a modeling prompt (see the next section). It was a gestural prompt
when coach McCall pointed to the place that Tom should stand in the batter’s
box. When coach McCall showed him the motion of the ball and where to

182 Chapter 10

swing the bat, he was using gestural prompts that helped Tom hit the ball. Consider
another example. A special education teacher shows a student two cards; one card
has the word EXIT and the other the word ENTER. The teacher then asks the stu-
dent to point to the word EXIT. Because the student does not know the word EXIT
(has never made the correct discrimination), the teacher provides a prompt to get
the student to point to the EXIT card: The teacher turns to look at the EXIT card.
If this gesture makes it more likely that the student will point to the EXIT card, it is
considered a gestural prompt. Gestural prompts can be effective because people
have a history of reinforcement for responding correctly to gestures. Therefore
gestural prompts have stimulus control over the behavior indicated by the gesture.

Modeling Prompts Any demonstration of the correct behavior by another person
that makes it more likely that the correct behavior will occur at the right time is a
modeling prompt. (Such a demonstration also is called modeling.) A person observes
the model and imitates the modeled behavior (makes the correct response) in the pres-
ence of the SD. When coach McCall hit the ball to show Matt how to do so, he was
modeling the correct behavior (providing a modeling prompt). Matt imitated the
coach’s behavior and hit the ball successfully himself. For a modeling prompt to be
successful, the person must be able to imitate the model’s behavior (Baer, Peterson, &
Sherman, 1967). Because imitation is a type of behavior that most people learn early
in life, most people benefit from observing models (Bandura, 1969). Modeling
prompts can be effective because people have a history of reinforcement for
imitating models, and as a result, develop a generalized imitative repertoire.
Therefore modeling prompts have stimulus control over imitative behavior.

Physical Prompts With a physical prompt, another person physically helps
the person to engage in the correct behavior at the right time. Coach McCall
held the bat with Trevor and physically helped him to swing the bat and hit the
ball. The person using a physical prompt is executing all or part of the behavior
with the learner. A physical prompt often involves hand-over-hand guidance, in
which the trainer guides the person’s hands through the behavior. For example,
an art teacher may guide a student’s hand when teaching how to mold clay. The
pitching coach moves the pitcher’s fingers into the correct position on the baseball
when teaching how to grip the ball to throw a certain type of pitch. In teaching
toothbrushing to a student with disabilities, the trainer puts a hand over the
student’s hand on the toothbrush and moves it in a brushing motion. In each of
these examples, when the person could not correctly execute the behavior with
the help of verbal, gestural, or modeling prompts, a physical prompt was used to
guide the person through the behavior. According to Sulzer-Azaroff and Mayer
(1991), physical prompts are appropriate when telling or showing the person the
behavior is ineffective (i.e., when verbal, gestural, and modeling prompts do not
evoke the behavior). Unless the person resists, most behaviors can be prompted
physically. (Language is an exception; you cannot physically prompt a person to
say something.) Physical prompting is also known as physical guidance.

All four types of response prompts involve the behavior of one person,
who tries to influence the behavior of another person (by issuing instructions,
modeling, and so on). Therefore, response prompts are intrusive; they involve
one person exerting control over another. In a teaching situation, this is necessary

Prompting and Transfer of Stimulus Control 183

and acceptable. However, you should always use the least intrusive type of response
prompt and resort to more intrusive prompts only when they are necessary to get the
person to engage in the appropriate behavior. As shown in Table 10-1, verbal
prompts are least intrusive and physical prompts are most intrusive.

Stimulus Prompts
A stimulus prompt involves some change in a stimulus, or the addition or removal
of a stimulus, to make a correct response more likely. A stimulus prompt might
involve a change in the SD or the S-delta (S�) that makes the SD more salient
(more noticeable or conspicuous) and the S� less salient so that the person is
more likely to respond to the SD (to make the correct discrimination). Likewise,
other stimuli might be used in conjunction with the SD or S� to make the SD

more salient, thereby making a correct discrimination more likely. Changing the
SD is called a within-stimulus prompt. Adding another stimulus or cue to the
SD is called an extrastimulus prompt (Schreibman, 1975).

Within-Stimulus Prompts You can change the salience of an SD (or S�) in a
number of ways. You can change the position of the SD or you can change some
dimension of the SD (or S�), such as size, shape, color, or intensity (Terrace,
1963a, b). Coach McCall used a stimulus prompt (in addition to response
prompts) when he taught his players to hit a baseball. The SD is a baseball
approaching the batter at normal speed from a normal distance. The response is
to swing the bat correctly, and the reinforcing consequence is hitting the ball and
getting praise from the coach.

How did coach McCall change the SD to make it easier for the kids to
hit the ball?

Coach McCall used a stimulus prompt when he had Dave throw easy pitches
at first (slow pitches, thrown a short distance from the batters, in easy locations to
hit). The easy pitch is a stimulus prompt: It is a change in the intensity of the SD

that makes it more likely that the kids could make the correct response and hit the
ball. The teacher who wanted the student to point to the EXIT sign would be
using a stimulus prompt if the EXIT sign were positioned closer to the student
than the ENTER sign (location), or if the EXIT sign were bigger than the
ENTER sign (size). Changing the size or location would make it more likely that
the student would point to the correct sign. The silver stripe down one strand of
the speaker wire in a stereo system is a stimulus prompt. This stripe makes it

TABLE 10-1 Ranking of Response Prompts by Level of Intrusiveness

Type of Response Prompt Level of Intrusiveness

Verbal Least (weakest)

Gestural Moderately low

Modeling Moderately high

Physical Most (strongest)

184 Chapter 10

more likely that you will put the two strands of the wire into the correct
connections on the stereo and the speakers. In each of these examples, the SD is
changed in some way to make it more likely that a correct response will occur
(within-stimulus prompt).

Extrastimulus Prompts Sometimes stimulus prompts involve adding a stimulus
to help a person make a correct discrimination (extrastimulus prompt). The line
drawn in the dirt next to home plate by the coach helps the T-ball player stand
in the correct place when batting. Wacker and Berg (1983) used picture prompts
to help adolescents with intellectual disabilities complete complex vocational
tasks correctly. The tasks involved assembling or packaging items. The picture
prompts helped the adolescents package or assemble the correct part at the
correct time. Alberto and Troutman (1986) recount an interesting use of a stim-
ulus prompt by a teacher who wanted to teach a group of young children to
identify their right hands. The teacher put an X on the back of each child’s
right hand to help the children make the correct discrimination. Over time, the
X wore off and the children continued to make the correct discrimination.
The gradual elimination of the X amounted to fading of the stimulus prompt
and transfer of stimulus control to the natural SD (the right hand). When a
student is learning multiplication facts using flash cards, the problem on the
flash card (e.g., 8 � 2) is the SD, and the answer on the opposite side of the
card is the stimulus prompt. It is an added stimulus that helps the student make
the correct response in the presence of the SD.

Types of Prompts

Response prompts: The behavior of another person evokes the correct response.

â–  Verbal prompts

â–  Gestural prompts

â–  Modeling prompts

â–  Physical prompts

Stimulus prompts: A change in some aspect of the SD or S� or the addition or removal of another
stimulus makes a correct response more likely.

â–  Within-stimulus prompts

â–  Extrastimulus prompts

Transfer of Stimulus Control

Once the correct response has occurred, the prompts must be eliminated, so as to
transfer stimulus control to the natural SD (Billingsley & Romer, 1983). Training
is not complete until Trevor can hit the baseball without any assistance, until
Natasha can read the words on the flash cards without a verbal prompt, and until
the young children can identify their right hands without the X. As these examples
suggest, the end result of transfer of stimulus control is that the correct behavior
occurs at the right time without any assistance (prompts).

Prompting and Transfer of Stimulus Control 185

There are a number of ways to transfer stimulus control: prompt fading,
prompt delay, and stimulus fading. The goal of each method is to move from
the artificial stimulus control of the prompts to the natural stimulus control of
the relevant SD.

Prompt Fading
Prompt fading is the most commonly used method of transferring stimulus control.
With prompt fading, a response prompt is removed gradually across learning trials
until the prompt is no longer provided (Martin & Pear, 1992). When coach McCall
provided fewer and fewer instructions to Luke as he hit the ball, the coach was fading
a verbal prompt. When the coach provided less and less physical guidance to Trevor as
he started to hit the ball successfully, the coach was fading a physical prompt.

How did the teacher fade verbal prompts when teaching Natasha to read words
on flash cards?

Initially, the teacher said the word as a verbal prompt; then she said part of
the word, then the first letter of the word, and eventually she didn’t say anything
after she presented the SD. Gradually saying less of the word over time was fading
the verbal prompt. In each of these examples, one type of prompt was faded; this
is known as fading within prompt. A study by Berkowitz, Sherry, and Davis (1971)
illustrated the use of physical prompting and fading within prompt to teach boys
with profound intellectual disabilities to eat with a spoon. Initially, the researchers
held the child’s hand with the spoon and physically prompted the entire behavior
of scooping food with the spoon and putting it in the mouth. They then faded the
physical prompt in seven steps until the child was using the spoon without
any assistance. Each fading step involved less and less physical assistance as the
physical prompt was eliminated gradually.

Transfer of Stimulus Control
â–  Prompt fading: The response prompt is eliminated gradually.

â–  Prompt delay: After the SD is presented, the prompt is delayed to provide the opportunity for an
unprompted response to occur.

â–  Stimulus fading: The stimulus prompt is eliminated gradually.

Sometimes you can eliminate a prompt in just one step. You might need to
tell a person only one time how to execute a behavior before the person performs
the correct behavior without another verbal prompt. Likewise, you might have
to model the behavior only once before the behavior occurs without further
prompting. It is also possible that, after only one physical prompt, the person
might engage in the correct behavior.

Another type of prompt fading involves fading across different types of
prompts or fading across prompts. Consider the following example. Lucy, a
woman with severe intellectual disability, works in the stockroom of the shoe
department at a major discount store. Her job is to take the paper stuffing out of
shoes so that the shoes can be displayed on the store shelves. She sits at a large

186 Chapter 10

table covered with shoes. (Another worker puts the shoes on the table.) After
she pulls the paper out of a pair of shoes, another worker moves the shoes to the
store shelves. The job coach has to teach Lucy how to do the job correctly. The
three-term contingency is as follows:

Because Lucy cannot execute the correct behavior, the job coach uses prompts to
get the behavior to occur and then fades the prompts. One method is least-to-most
prompting and fading. The job coach provides the least intrusive prompt first and
uses more intrusive prompts only as necessary to get the correct behavior to occur.
If Lucy does not pull the paper out of the shoes on her own, the job coach first says,
“Lucy, pull the paper out of the shoe.” This is the least intrusive verbal prompt.
If Lucy does not respond in 5 seconds, the job coach repeats the verbal prompt and
points at the paper in the shoe (provides a gestural prompt). If Lucy does not respond
in 5 seconds, the job coach models the correct behavior as she provides the verbal
prompt. If Lucy still does not respond, the job coach uses physical guidance as she pro-
vides the verbal prompt. She takes Lucy’s hand, helps her pull out the paper, and then
praises Lucy. On the next trial, the job coach goes through the same sequence until
Lucy responds correctly. Over trials, Lucy will make the correct response before the
physical prompt is needed, then before the modeling prompt, and then before the ges-
tural prompt, until eventually she needs no prompt at all to pull the paper out of the
shoe. The prompts were faded as Lucy needed less and less assistance. Least-to-most
prompting is used when the trainer believes the learner may not need a physical
prompt to engage in the correct behavior and wants to provide the opportunity for the
learner to perform the task with the least assistance necessary.

Another method of fading across prompts is most-to-least prompting and fading.
With this method, the most intrusive prompt is used first and is then faded to less intru-
sive prompts. Most-to-least prompting is used when the trainer believes the learner will
need a physical prompt to engage in the correct behavior. Using most-to-least prompt-
ing, the job coach would start by providing a physical prompt together with a
verbal prompt. She would then start to fade the physical prompt as Lucy successfully
executed the behavior. Once she faded the physical prompt, she would provide a
verbal and gestural prompt. Then, as Lucy continued to be successful, she would
fade the gestural prompt and provide only the verbal prompt. Finally, she would fade
the verbal prompt as Lucy correctly took the paper out of the shoe with no assistance.
Whether fading within prompt or across prompts, the ultimate goal is to transfer
stimulus control to the natural SD so that the prompts are no longer used.

Prompt Fading

Fading within prompt

Fading across prompts

â–  Least-to-most prompting

â–  Most-to-least prompting

Antecedent Behavior

Paper stuffing in the shoe (SD). Lucy pulls the stuffing out. The job coach praises Lucy.

Consequence

Prompting and Transfer of Stimulus Control 187

Prompt Delay
Another method to transfer stimulus control from a response prompt to the natural
SD is prompt delay. In this procedure, you present the SD, wait a certain number
of seconds, and then, if the correct response is not made, you provide the prompt.
The time delay between the presentation of the SD and the prompt may be
constant or progressive (Handen & Zane, 1987; Snell & Gast, 1981).

Cuvo and Klatt (1992) taught adolescents with disabilities to read common
words that they would encounter in everyday life (e.g., MEN, WOMEN, STOP,
ENTER). They used a constant prompt delay procedure: They presented a word
on a flash card (SD), and if the student did not respond in 4 seconds, they said
the word for the student (verbal prompt). The objective was for the student to
read the word within the 4 seconds before the prompt was provided. Eventually,
all students read the words within the 4 seconds and the prompts were no longer
used. Stimulus control was transferred from the verbal prompt to the written word.

Matson, Sevin, Fridley, and Love (1990) used a progressive or graduated
prompt delay procedure to teach children with autism to make appropriate social
responses (to say “Please,” “Thank you,” and “You’re welcome”). To teach a child
to say “Thank you,” the experimenter gave the child a toy (SD), and if the
child said “Thank you,” the experimenter delivered an edible reinforcer and praise.

However, because the children with autism did not say “Thank you,” the
trainer delivered a verbal prompt (he said “Thank you”) 2 seconds after giving the
toy to the child, and the child imitated the verbal prompt. These children had
already demonstrated the ability to imitate verbal prompts, so Matson knew that a
verbal prompt would evoke the correct behavior. After the child said “Thank you”
when the prompt delay was 2 seconds, it was gradually increased by 2-second inter-
vals until the prompt delay was 10 seconds. Eventually, as the prompt delay
increased from 2 to 10 seconds, the child started to say “Thank you” before the
prompt was given. Once this occurred consistently, the prompt was no longer
given because stimulus control had transferred to the natural SD (Figure 10-3).

Whether the prompt delay is constant or graduated, the first trial always begins
with a 0-second delay between the SD and the prompt. In subsequent trials, the
prompt delay is inserted to allow the person to make the correct response before
the prompt is given. If the person cannot make the correct response, the prompt is
provided to evoke the response in the presence of the SD. Eventually, after the
correct response is prompted and reinforced in a number of trials, the response will
occur after the SD is presented but before the prompt is delivered. Once this
happens consistently, stimulus control has been transferred from the prompt to
the SD.

Antecedent Behavior

Outcome: Child is more likely to say “Thank you” when receiving a toy from another person.

Child is given a toy (SD). Child says “Thank you.” Child receives edible reinforcer and praise.

Consequence

188 Chapter 10

Stimulus Fading
Whenever stimulus prompts are used to get a correct response to occur, some aspect
of the SD or the stimulus situation is changed to help the person make the correct
discrimination. Eventually the stimulus prompts must be removed through a process
of stimulus fading to transfer stimulus control to the natural SD. If the stimulus
prompt involved adding a stimulus to get the correct response to occur (extrastimulus
prompt), stimulus fading would involve gradually removing that additional stimulus
as the response began occurring reliably in the presence of the SD. Once this addi-
tional stimulus is completely removed and the response continues to occur in the
presence of the SD, stimulus control has been transferred to the SD. When students
are using flash cards to learn multiplication facts, the answer on the other side of the
flash card is a stimulus prompt. The students are using stimulus fading when they
look at the answers to the problems less and less as they go through the flash cards.

Pe
rc

en
ta

ge
o

f T
ar

ge
t B

eh
av

io
rs

Yo
u’

re
W

el
co

m
e

Th
an

k
Yo

u
Pl

ea
se

Baseline Treatment Package Follow-Up

Sessions

0

100

5 10 15 20 25 30 35 40 45 50

•
•••

•
•••••••••

•

•••
•

•

••
•••

•
••

•
••••••

•
•

•
•

•
•

•••••••
0

100

••
•

••••••••••• ••••••

•

•
•

•
•
••

•
•

•
•

•••
•

•
••

•
•

•

••

•••••
0

100

•

•
•

•

•
•

••••

••
•

•

•
•

••
•

•
••

•••••••••••
•

••••••••••••••

Untrained
Stimuli

Target
Stimuli

Child 2

FIGURE 10-3 This graph shows the acquisition of three social behaviors in a child with autism after the use of a
graduated prompt delay procedure. This graph illustrates a multiple-baseline-across-behaviors
research design. (From Matson, J. L., Sevin, J. A., Fridley, D., & Love, S.R. (1990). Increasing spon-
taneous language in three autistic children. Journal of Applied Behaviour Analysis, 23, 227–233.
Copyright © 1990 Society for the Experimental Analysis of Behavior. Reprinted by permission of the
Society for the Experimental Analysis of Behavior.)

Prompting and Transfer of Stimulus Control 189

Once they get all the problems correct and no longer look at the answers, stimulus
control has transferred from the written answers (stimulus prompt) to the problems
(the SD). When a young child has an X on the back of her right hand, this stimulus
prompt helps her identify her right hand. As the X wears off over the course of a few
days, stimulus fading occurs. When the child identifies her right hand without the X,
stimulus control has been transferred from the prompt to the natural SD.

Stimulus fading is also used when the stimulus prompt involves a change in
some aspect of the SD itself (within-stimulus prompt). In this case, stimulus fading
would involve gradually changing the SD from its altered form to its natural form.
Coach McCall used a stimulus prompt when he had Dave throw easy pitches for
the players to hit. In this case, stimulus fading involved gradually increasing the
distance and speed of the pitches until they were being thrown at normal distance
and speed. Gradually increasing the distance and speed while the kids continued
to hit the ball successfully corresponded to fading the stimulus prompt and
transferring stimulus control to the natural SD (a pitch of normal speed).

The teacher who was teaching the student to point to the EXIT sign used a
stimulus prompt by making the EXIT sign bigger than the ENTER sign.

How would the teacher use stimulus fading?

The teacher would use stimulus fading by reducing the size of the EXIT sign
gradually until it was the same size as the ENTER sign. Once they were the
same size, the stimulus prompt would be gone and stimulus control would be
transferred from the increased size of the word (prompt) to the word itself (SD).

Note that some authors distinguish between stimulus fading and stimulus
shaping (Cooper et al., 1987; Etzel, LeBlanc, Schilmoeller, & Stella, 1981).
Although there is a technical difference between the two procedures, they are
quite similar (Deitz & Malone, 1985), and both involve the gradual removal of
a stimulus prompt to transfer stimulus control. For this reason, and to avoid confu-
sion between shaping (see Chapter 9) and stimulus shaping, the term stimulus
fading is used here to refer to all procedures involving gradual removal of a stimu-
lus prompt. (To study the distinction between stimulus fading and stimulus
shaping, see Cooper et al., 1987; Etzel & LeBlanc, 1979; or Etzel et al., 1981.)

FOR FURTHER READING
Varied Applications of Prompting and Fading

Prompting and fading have been used widely in applied behavior analysis to teach a variety of skills in a variety
of populations of learners. One area in which prompting and fading is used widely is in teaching skills to chil-
dren with autism. For example, a number of authors have shown that written scripts could be used as prompts
to help children with autism initiate social interaction. The scripts are then faded as the children continue to
exhibit appropriate social behavior (e.g., Krantz & McClannahan, 1993, 1998; Sarokoff, Taylor, & Poulson,
2001). Another application of prompting and fading is in the area of staff management. In a study by Petscher
and Bailey (2006), staff members in a classroom for students with disabilities were prompted to engage in
specific instructional activities via a vibrating pager they carried with them. If the staff did not engage in an
instructional activity at the appropriate time, their pager vibrated as a prompt for them to do so. Once they
began engaging in the correct behavior at the correct time, they continued to do so even when the prompts
were discontinued. Another area of application of prompting and fading is with athletic skills. For example,
Osborne, Rudrud, and Zezoney (1990) used stimulus prompts to enhance baseball players’ ability to hit curve-
balls. In another example, Luyben, Funk, Morgan, Clark, and Delulio (1986) used prompting and fading to
improve soccer passing skills of individuals with severe intellectual disabilities.

190 Chapter 10

How to Use Prompting and Transfer
of Stimulus Control

When your objective is to develop appropriate stimulus control over a behavior
(to ensure that a new or existing behavior occurs in the correct circumstances
at the correct time), you will use prompting and transfer of stimulus control.
Before deciding to use such procedures, it is important to determine whether
you are addressing a problem of stimulus control or a problem of noncompli-
ance (a “can’t do” problem or a “won’t do” problem). If the person has
not learned the behavior or has not learned to do the behavior in the correct
situation (“can’t do”), the appropriate procedure is prompting and transfer of
stimulus control. However, if the person has exhibited the correct behavior in
the correct situation in the past but now refuses to do it (“won’t do”), the prob-
lem is noncompliance, and prompting and transfer of stimulus control would
not be the most appropriate procedure. See Chapters 13-19 for procedures to
treat noncompliance and other behavior problems. The following guidelines
should be observed in prompting and transferring stimulus control (see also
Alberto & Troutman, 1986; Martin & Pear, 1992; Rusch, Rose, & Greenwood,
1988; Sulzer-Azaroff & Mayer, 1991).

1. Choose the most appropriate prompting strategy. A variety of response
prompts and stimulus prompts is available. You need to choose the one
that best fits the learner and the learning task. If a novel behavior is
being taught, response prompts are most appropriate because they can be
used to generate a new behavior in the appropriate situation. For learners
with limited abilities (e.g., people with developmental disabilities or young
children), stronger or more intrusive prompts such as physical prompts are
most appropriate. Less intrusive or weaker prompts (such as verbal
prompts) should be used if the learner is capable of benefiting from them
(i.e., the learner is capable of following instructions). If unsure of the level
of prompt that is needed, you may use graduated prompt strategies, such
as the least-to-most strategy (also called the system of least prompts), in
which less intrusive prompts are tried first and more intrusive prompts are
used as needed. Stimulus prompts are most appropriate when you want to
help a person make a correct discrimination. Because stimulus prompts
highlight the SD (make it more salient), they increase the likelihood that
the learner will respond when the SD is present.

2. Get the learner’s attention. Before you present the instructional stimuli
(the SD or the prompts), be sure that the learner is paying attention.
Reduce or eliminate distractions and competing stimuli and, if necessary,
prompt and reinforce the learner’s attention before beginning an instruc-
tional trial. For example, to get Matt’s attention before he provides a model-
ing prompt, coach McCall might say, “Matt, watch how I swing the bat.”

3. Present the SD. The learning trial always starts with the presentation of the
SD. This is the stimulus that should evoke the correct response in the
learner once training is completed. If the learner makes the correct response
in the presence of the SD, prompts are not necessary. There may be some
exceptions in which a verbal or modeling prompt comes before the SD,

Prompting and Transfer of Stimulus Control 191

such as the coach providing instructions or modeling how to swing the bat
before the pitch was thrown (the SD). However, this example is an excep-
tion; in most cases, the learning trial starts with the presentation of the SD.

4. Prompt the correct response. If the SD does not evoke the correct response, the
prompt should be provided. When using a stimulus prompt, you will change
the stimulus situation in some way when presenting the SD or you will
change some aspect of the SD. When using a response prompt, you will
present the SD and then immediately deliver the appropriate prompt.

5. Reinforce the correct behavior. When the learner engages in the correct
behavior (whether prompted or unprompted) in the presence of the SD,
immediately provide a reinforcer. Because the goal is for the learner to
engage in the correct behavior without prompts when the SD is present,
you should increase the magnitude of reinforcement for unprompted
responses. For example, praise should be more enthusiastic or a larger
amount of a reinforcer should be given.

6. Transfer stimulus control. As soon as possible, prompts should be eliminated
to transfer stimulus control from the prompt to the natural SD. If response
prompts are being used, fading or prompt delay procedures can be used to
transfer stimulus control. If stimulus prompts are being used, transfer stimu-
lus control by means of stimulus fading procedures. When fading response
or stimulus prompts, the fading steps should be small (i.e., the process should
be gradual) so that the person continues to engage in the correct behavior as
the prompts are faded. If a fading step is too big, the correct behavior may be
lost (errors may occur). If this happens, you should back up to a previous
fading step and provide more of the prompt or a stronger (more intrusive)
prompt. When using a prompt delay procedure, you can enhance transfer
of stimulus control by providing more reinforcement for responses that
occur during the delay before the prompt is given.

7. Continue to reinforce unprompted responses. When the correct behavior is
occurring in the presence of the SD after prompts have been eliminated,
continue to reinforce the behavior. As the learner continues to engage in
the correct behavior, switch from a continuous reinforcement schedule to
an intermittent reinforcement schedule. This will help to maintain the
correct behavior over time. The goal is for the behavior to eventually come
under the control of natural contingencies of reinforcement. For example,
once Luke learns to hit the baseball, getting a hit is a natural reinforcer.

Guidelines for Prompting and Transferring Stimulus Control

1. Choose the most appropriate prompting strategy.

2. Get the learner’s attention.

3. Present the SD.

4. Prompt the correct response.

5. Reinforce the correct behavior.

6. Transfer stimulus control by fading or prompt delay.

7. Continue to reinforce unprompted responses.

192 Chapter 10

Prompting and Transfer of Stimulus Control
in Autism Treatment

One common application of prompting and transfer of stimulus control is
teaching skills to children with autism. Children with autism often receive early
intensive behavioral intervention (EIBI) during which behavior analysts or other
behaviorally trained professionals teach important academic skills so that the
children with autism can develop the same skills as their peers without disabilities
and achieve success when they enter school. Before conducting EIBI the behavior
analyst conducts an assessment and identifies the sequence of skills that need to
be trained. Prompting and fading (transfer of stimulus control) procedures are
then used to train each skill in sequence. For example, a beginning sequence of
skills for a young child with autism might include: (1) making eye contact, (2)
imitating gross motor movements, (3) imitating actions with objects, (4) following
simple instructions, and so on (see Taylor & McDonough, 1996 for an example of
a curriculum for a young child with autism). Each of these skills is important for
learning other skills; more advanced skills build on these basic skills.

Let’s consider how prompting and fading would be used to teach the skill of
imitating a gross motor movement. You would start by having the child sit across a
small table from you with no other distractions. After getting the child’s attention,
you present the SD (say “do this” while clapping your hands), prompt the correct
response (physically prompt the child to clap by taking his hands in yours and
clapping them), and immediately provide a reinforcer such as praise and a small
edible. This sequence of presenting the SD, prompting the response, and provid-
ing a reinforcer is called a learning trial. You repeat the learning trial a number
of times and, in each subsequent trial, you provide less and less physical prompt-
ing (fading) until the child is clapping independently when you say “do this” and
clap your hands. Once the child imitates this motor movement without prompts,
you choose another movement (e.g., tap the table, wave, place arms up, etc.) to
teach. Using the same sequence of presenting the SD, prompting the response,
reinforcing the response, and then fading the prompts over trials until the imita-
tive response occurs without prompts, you teach the child to imitate a number of
different motor movements when you say “do this” and make the movement.
Eventually, the child will imitate any motor movement you make when you say
“do this” and make a movement. At this point you can say the child has learned
to imitate gross motor movements and you can move on to the next skill in the
sequence. Depending on the ability level of the child it might take days or weeks
until the child has developed the skills to imitate movements and you can go on
to the next skill. Taking data on the percentage of correct responses across trials
will tell you when the child has mastered the skill.

CHAPTER SUMMARY

1. A prompt is the behavior of another person deliv-
ered after the presentation of the SD, a change in
the SD, or the addition of a stimulus with the SD.

Prompts are used to increase the likelihood that a
correct behavior will occur in the correct situation
(in the presence of the SD).

Prompting and Transfer of Stimulus Control 193

2. Fading is the gradual elimination of a prompt.
Fading is used to get the behavior to occur in
the presence of the SD without any prompts.

3. Response prompts occur when the learner’s
behavior is evoked by the behavior of another per-
son. Stimulus prompts involve a change in some
aspect of the SD or some other stimulus change
that makes a correct discrimination more likely.

4. Response prompts include verbal prompts, ges-
tural prompts, physical prompts, and modeling.

5. Transfer of stimulus control is the elimination of
the prompt to get the behavior under the stimulus
control of the relevant SD. Transfer of stimulus
control procedures involve fading and prompt
delay. In fading, a response prompt or a stimulus
prompt is eliminated gradually until the response
occurs in the presence of the SD without any
prompt. In a prompt delay procedure, a period
elapses between the presentation of the SD and
the delivery of the response prompt.

KEY TERMS

extrastimulus prompt, 184
fading, 180
gestural prompt, 179
modeling prompt, 179
physical guidance, 183
physical prompt, 179

prompt, 178
prompt delay, 188
prompt fading, 186
response prompt, 182
stimulus fading, 189
stimulus prompt, 182

transfer of stimulus control, 185
verbal prompt, 178
within-stimulus prompt, 184

PRACTICE TEST

1. What is a prompt? When is a prompt used in
behavior modification? (pp. 178–179)

2. What is a response prompt? Identify and describe
four types of response prompts. (pp. 182–184)

3. Provide examples of the four types of response
prompts. (pp. 182–184)

4. What is a stimulus prompt? Describe two types
of stimulus prompts. (pp. 184–185)

5. Provide examples of the two types of stimulus
prompts. (pp. 184–185)

6. What is least-to-most prompting? What is another
term for it? Provide an example. (p. 187)

7. What is most-to-least prompting? Provide an
example. (p. 187)

8. Flashing lights on a billboard that make it more
likely you will read the words on the billboard
are a kind of prompt. What prompt is used in
this example? (pp. 184–185)

9. What is transfer of stimulus control? Why is it
important? (pp. 185–186)

10. Describe fading of response prompts. Provide an
example. (pp. 186–187)

11. Describe fading with least-to-most prompting and
fading with most-to-least prompting. (p. 187)

12. Describe fading of stimulus prompts. Provide an
example of within-stimulus prompt fading and
extrastimulus prompt fading. (pp. 189–190)

13. Describe the prompt delay procedure. Provide
an example of the constant prompt delay proce-
dure and an example of the progressive prompt
delay procedure. (p. 188)

14. Suppose you are conducting a learning trial with
a student with autism. How would you use a
verbal and physical response prompt to get the
student to pay attention to you when you said
the student’s name (SD)? (p. 193)

15. Describe how you could use stimulus prompts
and fading to learn definitions for the behavior
modification procedures described in this
chapter. (p. 189)

194 Chapter 10

APPLICATIONS

1. Describe how you would use prompting and
fading to teach your 6-month-old puppy to
come to you at the command “Come.” Assume
that you have a 20-foot leash and a pocketful of
bite-size dog treats to use during training.

2. You are interested in playing golf, but your put-
ting is so bad that you are too embarrassed to
play with any of your friends. You decide that
you will use stimulus prompting and fading to
improve your putting. Let’s suppose that you
have access to a putting green. Describe three

different ways you could use stimulus prompt-
ing and fading to improve your putting. Be
creative and assume that you can manipulate
the putter, the green, the golf ball, or the hole
in any way you want.

3. Your 16-year-old niece, Edie, has been pestering
you to teach her how to drive your car. You
finally give in and take her to an empty mall
parking lot for her first lesson. Describe how
you will use most-to-least prompting and fading
to teach her how to drive.

MISAPPLICATIONS

1. Little Gloria was just starting to babble and
make some recognizable word sounds. Her par-
ents were thrilled. Her father, who had taken a
behavior modification class, decided that he
would use prompting and fading to get Gloria
to say “Mama” and “Dada.” What is wrong with
this application of prompting and fading? What
would be a more appropriate behavioral proce-
dure to use with Gloria?

2. Every day, it is Roger’s job to set the table for
supper. Although Roger had set the table every
day for weeks, he recently became interested in
the TV game show Jeopardy!, which airs at the
time that he is supposed to set the table. Now
Roger watches Jeopardy! when he should be set-
ting the table. His father reminds Roger each
day, but he ignores the request and continues
to watch TV. His father has decided he will

use prompting and fading to get Roger to set
the table. What is wrong with this application
of prompting and fading? What would be a
better procedure to use with Roger?

3. Michelle is a youngster with autism. She types
with physical prompts from her teacher, who
holds Michelle’s hand as she types. Michelle
has been typing words and sentences to commu-
nicate on the keyboard for more than a year and
her teacher continues to provide the physical
prompts as she types. If the teacher does not
have her hand over Michelle’s hand, Michelle
will not type any words; therefore, her teacher
continues to provide the physical prompts, and
Michelle continues to communicate on the key-
board. What is wrong with this application of
prompting and transfer of stimulus control?

Prompting and Transfer of Stimulus Control 195

This page intentionally left blank

Chapter Eleven

Chaining

As we have seen, prompts are used to evoke a behavior, and transfer of stimuluscontrol is used to eliminate the prompts and get the behavior to occur in the
presence of the relevant discriminative stimulus (SD). Most often, these proce-
dures are used to develop simple discriminations, in which one response occurs in
the presence of one SD. For example, a baseball player swings the bat to hit a

baseball. A student reads a word correctly. You plug the speaker
wire into the right outlet. You say “Thank you” when someone
gives you something. Each of these examples involves one behavior
occurring in the correct situation. However, many situations call for
complex behaviors that have multiple component responses. A com-
plex behavior consisting of many component behaviors that occur
together in a sequence is called a behavioral chain.

Examples of Behavioral Chains

When you want a piece of chewing gum, you have to engage in a
sequence of responses. You perform the following actions: (1) reach
into your pocket, (2) pull out the pack of gum, (3) pull a single stick
out of the pack, (4) unwrap the piece of gum, and (5) put the gum into
your mouth. Getting a piece of gum involves at least five behaviors,

which must occur together in the correct sequence. You can engage in a particular
behavior in the sequence only if the previous behavior in the sequence has been com-
pleted. You can’t put the gum in your mouth unless you have unwrapped it. (Actually,
you could, but why would you want to?) You can’t unwrap the gum unless you have
pulled a piece out of the pack. You can’t pull a piece of gum out of the pack unless
you have gotten the pack of gum out of your pocket.

Consider another example. Bobby works for an industrial laundry company.
Her job is to fold towels and put them into boxes so they can be shipped to the
customers (e.g., hotels, health clubs, hospitals). As they come out of the dryer,
another worker brings the towels over to Bobby in a big bin. Bobby’s job consists
of the following behavioral chain: (1) she grabs a towel from the bin, (2) lays it
out flat on the table, (3) grabs one end and folds it in half, (4) grabs one end
of the half-folded towel and folds it in half again, (5) grabs one end of the

â–  What is a stimulus-response chain?

â–  Why is it important to conduct a task
analysis of a stimulus-response chain?

â–  How do you use forward chaining and
backward chaining to teach a chain of
behaviors?

â–  What is total task presentation, and
how does it differ from the chaining
procedures?

â–  What are three other strategies for
teaching behavioral chains?

197

quarter-folded towel and folds it in half again, (6) picks up the folded towel, and
(7) puts it into the box. When the box is full, another worker loads the box of
towels on a truck. Bobby’s job of folding towels consists of a seven-step behavioral
chain. Each behavior in the chain can be completed only after the previous beha-
viors in the chain have been completed in sequence. Each component behavior
in the chain depends on the occurrence of the previous behavior.

This chapter describes how to analyze the components of a behavioral chain and
how to use various methods to teach a person to engage in a chain of behaviors.

Analyzing Stimulus–Response Chains

Each behavioral chain consists of a number of individual stimulus-response compo-
nents that occur together in a sequence. For this reason, a behavioral chain is often
called a stimulus-response chain. Each behavior or response in the chain produces
a stimulus change that acts as an SD for the next response in the chain. The
first response produces an SD for the second response in the sequence. The second
response produces an SD for the third response in the sequence, and so on, until all
the responses in the chain occur in order. Of course, the whole stimulus-response
chain is under stimulus control, so the first response in the chain occurs when a partic-
ular SD is presented. The gum in your pocket is an SD for the first response in the
chain—reaching in your pocket and grabbing the pack of gum. A bin full of towels
near Bobby is an SD for Bobby’s first response—grabbing a towel from the bin. Of
course, a behavioral chain continues only if the last response in the chain results in a
reinforcing consequence. Chewing the gum is a reinforcer for the behavioral chain of
putting the gum in your mouth. The folded towel in the box is a conditioned rein-
forcer for the behavioral chain of folding the towel. The folded towel is a conditioned
reinforcer because it is associated with other reinforcers, such as getting paid and being
praised by the boss.

The sequence of stimulus and response components involved in the behav-
ioral chain of getting a piece of gum is as follows:

1. SD1 (pack of gum in your pocket) � R1 (reach into your pocket)
2. SD2 (your hand in your pocket) � R2 (pull out the pack of gum)
3. SD3 (pack of gum in your hand) � R3 (pull out one stick of gum)
4. SD4 (one stick of gum in your hand) � R4 (unwrap stick of gum)
5. SD5 (unwrapped stick of gum in your hand) � R5 (put the gum in your

mouth) reinforcer (chewing the gum)

As you can see, each response creates the stimulus situation that is the SD for
the next response. Therefore, the next response in the chain depends on the
occurrence of the previous response.

A five-component stimulus-response chain can be illustrated in the following way:

SD1 � R1
SD2 � R2

SD3 � R3
SD4 � R4

SD5 � R5 � reinforcer

198 Chapter 11

Analyze the seven stimulus-response components involved in Bobby’s job of fold-
ing a towel and putting it into the box.

1. SD1 (a bin full of towels) � R1 (grab a towel from the bin)
2. SD2 (towel in hand) � R2 (lay towel flat on the table)
3. SD3 (towel flat on the table) � R3 (fold towel in half)
4. SD4 (half-folded towel on table) � R4 (fold towel in half again)
5. SD5 (quarter-folded towel on table) � R5 (fold towel in half again)
6. SD6 (folded towel on table) � R6 (pick up folded towel)
7. SD7 (folded towel in hand) � R7 (place towel in box) � reinforcer

(folded towel in box)

Once another worker brings a bin of towels over to Bobby, the full bin is the
first SD that has stimulus control over the first response in the stimulus-response
chain. Each subsequent response in the chain occurs because the previous
response created the SD that has stimulus control over that response.

Before we proceed, let’s look more closely at the beginning of the stimulus–
response chain. We can make the outcome of the chain more reinforcing by
means of an establishing operation. In our first example, the establishing opera-
tion makes the gum more reinforcing at a certain time, and this increases the
likelihood that you will start the behavioral chain by reaching into your pocket
and grabbing the pack of gum. The establishing operation might be having a
bad taste in your mouth from onions, having an old piece of gum in your
mouth, having just smoked a cigarette, or any circumstance that would make
fresh breath reinforcing at the time (such as talking to your girlfriend or boy-
friend). In this situation, you might say that you “want gum,” but that statement
does not help us understand why gum might be more reinforcing at a particular
time. It is better to look for stimuli or events that may function as establishing
operations.

Task Analysis

The process of analyzing a behavioral chain by breaking it down into its individ-
ual stimulus-response components is called a task analysis. Any time your goal is
to teach a complex task involving two or more component responses (a behavioral
chain) to a person, the first step is to identify all the behaviors that are necessary to
perform the task and write them down in order. Next, you identify the SD associ-
ated with each behavior in the task. Because teaching the task to the person
involves discrimination training with each stimulus-response component of the
behavioral chain, you must have a detailed task analysis that gives you an accurate
understanding of each stimulus-response component.

A task analysis to identify the right sequence of behaviors in a chain may be
conducted in various ways (Cooper, Heron, & Heward, 1987; Rusch, Rose, &
Greenwood, 1988). One way is to observe a person engage in the task and record
each of the stimulus-response components. For example, Horner and Keilitz
(1975) conducted a study in which they taught adolescents with intellectual

Chaining 199

disabilities to brush their teeth. The authors developed a task analysis of tooth-
brushing by observing staff members brush their teeth. Another method is to ask
a person who performs the task well (an expert) to explain all the components in
the task. Finally, you can develop a task analysis by performing the task yourself
and recording the sequence of responses in the task. Bellamy, Horner, and
Inman (1979) suggest that the advantage of performing the task yourself when
developing a task analysis is that it provides the best information about each
response involved in the task and the stimulus associated with each response.
That is, you can get the most information on a task from your own experience
with the task.

Different Ways to Conduct a Task Analysis
â–  Observe a competent person engage in the task.

â–  Ask an expert (a person who performs the task well).

â–  Perform the task yourself and record each of the component responses.

Once you have developed your initial task analysis, you might have to revise it
after you start training. You might find that you can break some behaviors down
into component behaviors, or that you can combine two or more behaviors into a
single behavior. Whether you revise your task analysis depends on how well your
training is progressing. If the learner is having difficulty with a certain behavior in
the chain, it might help to break down the behavior into two or more component
behaviors. However, if the learner can master larger units of behavior, two or
more component behaviors can be combined into one. Consider the following
example.

You want to teach a child with profound intellectual disability to eat with a
spoon. You have established the following task analysis.

1. SD1 (bowl of food and spoon on the table) � R1 (pick up the spoon)
2. SD2 (spoon in hand) � R2 (put spoon into food in the bowl)
3. SD3 (spoon in the food) � R3 (scoop food onto the spoon)
4. SD4 (food on the spoon) � R4 (lift spoonful of food from the bowl)
5. SD5 (holding spoonful of food) � R5 (put the food into the mouth) �

reinforcer (eat the food)

There are five steps or components to this task analysis. Each step consists of a
stimulus (SD) and response. This task analysis might be ideal for some children
learning how to eat with a spoon. However, for people who can more easily
master larger steps, you might want to combine some steps. The task analysis
with some combined steps might be as follows.

1. SD1 (bowl of food and spoon on table) � R1 (pick up spoon and put it
into the food in the bowl)

2. SD2 (spoon in the food) � R2 (scoop food onto the spoon)
3. SD3 (food on the spoon) � R3 (lift the spoonful of food and put it into

the mouth) � reinforcer (eat the food)

200 Chapter 11

As you can see, the only difference between this three-step task analysis and
the five-step task analysis is that the five-step task analysis breaks down the behavior
into smaller units. Each step is still characterized by a stimulus (SD) and a
response, but the size of the response is different. For some learners, the five-step
task analysis might be more appropriate; for others, the three-step task analysis
might be more appropriate. There is no right or wrong number of steps in a task
analysis. The only way to determine whether you have the correct number of
steps is to determine how well the task analysis works for a particular learner.

In a number of studies, researchers have developed task analyses of complex
tasks and then trained subjects to engage in the tasks. For example, Cuvo, Leaf,
and Borakove (1978) developed a task analysis for each of six janitorial skills,
which they then taught to people with intellectual disabilities. There were 13–56
steps in the task analyses of the six skills. Alavosius and Sulzer-Azeroff (1986)
taught staff in a treatment facility how to safely lift and transfer residents with
physical disabilities out of their wheelchairs. They developed an 18-step task anal-
ysis of the lift-and-transfer task. Other complex skills that have been subjected to

1
2
3
4
5
6
7
8
9

10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25

Parts in bin
Bearing on table
Nut in one corner
Nuts in two corners
Nuts in three corners
Cam in bearing
Roller in bearing
Red spring placed
Bearing rotated
Roller in bearing
Green spring placed
Bearing cleaned
Bearing in bag

Pick up bearing and place on table
Place hex nut in one bearing corner
Place hex nut in second corner
Place hex nut in third corner
Place cam base in bearing
Place roller in bearing
Place red spring in bearing
Rotate bearing and cam 180
Place roller in bearing
Place green spring in bearing
Wipe bearing with cloth
Place bearing in bag
Place bag in box

D
S Response Successive Trials

FIGURE 11-1 This task analysis data sheet has two columns to list the discriminative stimulus (SD) and response
for each component in the chain. Researchers use this data sheet to record progress when using a
chaining procedure to train a person in a complex task. (From Bellamy, G. T., Horner, R. H., &
Inman, D. P. [1979]. Vocational habilitation of severely retarded adults. Austin, TX: Pro-Ed Journals.
Reprinted by permission of the author.)

Chaining 201

task analysis include menstrual care skills (Richman, Reiss, Bauman, & Bailey,
1984), apartment upkeep skills (Williams & Cuvo, 1986), pedestrian skills for
walking safely through traffic (Page, Iwata, & Neef, 1976), leisure skills (Schleien,
Wehman, & Kiernan, 1981), and the skills college students need to write instruc-
tional manuals for community volunteers (Fawcett & Fletcher, 1977). Figure 11-1
shows a task analysis data sheet that might be used to record the learner’s progress
on a complex task. Note that the data sheet lists all of the SDs and responses in
the task. The number on the right corresponding to each task step is circled
when the learner masters that step (i.e., can complete the step without prompts).

Once the task analysis of a complex skill has been developed, the next step is
to choose a strategy for teaching the skill. Strategies for teaching complex tasks
(behavioral chains) are called chaining procedures. Chaining procedures
involve the systematic application of prompting and fading strategies to each
stimulus-response component in the chain. Three different chaining procedures
are described in the following sections: backward chaining, forward chaining, and
total task presentation.

Backward Chaining

Backward chaining is an intensive training procedure typically used with learners
with limited abilities. With backward chaining, you use prompting and fading to
teach the last behavior in the chain first. By starting with the last behavior in the
chain, the learner completes the chain on every learning trial. Once the last behavior
is mastered (once the learner exhibits the behavior on presentation of the SD, without
any prompts), you teach the next to last behavior. Once this behavior is mastered and
the learner engages in the last two behaviors in the chain without any prompts, the
next behavior up the chain is taught. This continues until the learner can exhibit
the whole chain of behaviors when presented with the first SD, without any prompts.
As an example, consider the use of backward chaining to teach Jerry, a young man
with severe intellectual disability, how to throw a dart at a dartboard. The task analysis
(adapted from Schleien et al., 1981) includes the following components.

1. SD1 (staff member says, “Jerry, let’s play darts”) � R1 (Jerry walks over to
the dartboard)

2. SD2 (standing near a line on the floor 8 feet from the dartboard) � R2
(Jerry walks up to the line and stands facing the dartboard with his toes
touching the line)

3. SD3 (standing at the line with a dart on an adjacent table) � R3 (Jerry
grasps dart between thumb and first finger, with the point facing the board)

4. SD4 (standing at the line and holding dart between thumb and first finger)
� R4 (Jerry bends his elbow, so that the forearm is at a 90-degree angle)

5. SD5 (standing at the line with dart in hand and elbow bent) � R5 (Jerry
thrusts forearm and hand toward the board and releases dart when arm is
extended) � reinforcer (dart hits the board)

To start the backward chaining procedure, you present the last SD (SD5),
prompt the correct response, and provide a reinforcer.

202 Chapter 11

SD5 + prompt � R5 reinforcer

In this example, you take Jerry over to the dartboard, prompt him to put his
toes up to the line, put the dart in his hand, and bend his elbow until his forearm
is at a 90-degree angle. This position is the SD for the last step in the chain (SD5).
Now you physically prompt the correct response. You hold Jerry’s hand in your
hand, thrust his hand forward, and release the dart when his arm is extended. As
the dart hits the dartboard, you praise Jerry. (Praise is a reinforcer for Jerry.) You
continue to physically prompt this response across learning trials, and as Jerry starts
to make the response himself, you begin to fade the prompt. You give him less
and less help, until he is throwing the dart himself as soon as you put the dart in
his hand and bend his elbow. Gestural prompts or modeling prompts may be used
instead of physical prompts, if these prompts have stimulus control over Jerry’s
behavior. You always use the least intrusive prompt necessary to get the behavior
to occur. Once Jerry has mastered the fifth component in the chain (once he
throws the dart independently as soon as you put it in his hand and bend his
elbow), you back up the chain and teach the fourth component.

To teach the fourth step in the chain, you arrange SD4, prompt the correct
response (R4), and provide praise as a reinforcer. You arrange SD4 by putting the
dart in Jerry’s hand as he is standing at the line. Once the dart is in his hand, you
physically prompt him to bend his elbow (R4). Once his elbow is bent (SD5),
Jerry will throw the dart (R5) because he has already learned to throw the dart
when he is holding the dart in his hand with his elbow bent. That is, throwing
the dart (R5) is already under stimulus control of SD5.

SD4 þ prompt � R4 � praise
SD5 � R5 � reinforcer

You fade your prompt by giving Jerry less assistance to bend his elbow until
he bends his elbow independently (without any prompts) as soon as SD4 is
presented. Now he has mastered the fourth and fifth components of the chain
and it is time to teach him the third component.

To teach the third component in the chain, you present SD3, prompt the
correct response (R3), and provide praise. You present SD3 by having Jerry stand
with his toes touching the line. Then you physically prompt him to pick up the
dart between his thumb and first finger (R3). Once the dart is in his hand (SD4),
Jerry will bend his elbow (R4) and throw the dart (R5) because he has already
learned these behaviors. (They are already under stimulus control of SD4.)

SD3 þ prompt � R3 � praise
SD4 � R4

SD5 � R5 � reinforcer

You fade your physical prompt, and as Jerry gets less assistance, he starts to
pick up the dart on his own. Once Jerry picks up the dart without any prompts as
soon as he is brought to the line, he has mastered this step. (R3 is under stimulus
control of SD3.) Now it is time to teach him the second step in the chain.

To teach the second step, you present SD2, prompt the correct response (R2),
and provide praise. You present SD2 by bringing Jerry to the side of the room where

Chaining 203

the dart-board is located, and then physically prompt him to step up to the line
(R2). Once Jerry is standing at the line (SD3), he will pick up a dart (R3), bend
his elbow (R4), and throw the dart at the dartboard (R5). He has already learned
the last three behaviors, so he will execute them as soon as the relevant SD is
presented.

SD2 þ prompt � R2 � praise
SD3 � R3

SD4 � R4
SD5 � R5 � reinforcer

As you fade your prompts, Jerry will walk up to the line without assistance
when he is presented with S2. Now it is time to teach the first step in the chain.

To teach the first step, you present SD1 (you say, “Jerry, let’s play darts”),
prompt response R1 (walking to the side of the room where the dartboard is
located), and provide praise. Once Jerry walks over to the side of the room where
the dartboard is located, he will then walk up to the line, pick up a dart, bend his
elbow, and throw the dart, because these four behaviors are under the stimulus
control of SD2 (being near the dartboard), and SD2 is the outcome of R1, the
behavior you are prompting.

SD1 þ prompt � R1 � praise
SD2 � R2

SD3 � R3
SD4 � R4

SD5 � R5 � reinforcer

Once you fade the prompts, Jerry will walk over to the dartboard indepen-
dently as soon as you say, “Jerry, let’s play darts” (SD1). Now the whole chain of
behaviors is under the stimulus control of SD1. As soon as you say, “Jerry, let’s
play darts,” he will be able to walk over to the dartboard, step up to the line, pick
up a dart, bend his elbow, and throw the dart.

In backward chaining with Jerry, each trial ended with the dart hitting the
board. Because you praised him each time the dart hit the dartboard, the dart
hitting the board is now a conditioned reinforcer for throwing the dart. Also,
because you praised him each time he engaged in the behavior at each training
step, each SD generated by the behavior is also a conditioned reinforcer. For exam-
ple, because you praised Jerry when he walked up to the line, standing at the line
was associated with praise and, therefore, was established as a conditioned rein-
forcer. Because you praised him for picking up the dart, holding the dart in his
hand is now a conditioned reinforcer. As you can see, using reinforcers at each
step in the backward chaining process is important because it makes the outcome
of each step a conditioned reinforcer, as well as an SD for the next response.

After Jerry is playing darts independently, you can start to praise him intermit-
tently to help maintain the behavior. Also, you can start to praise him when he
gets more points on the dartboard to reinforce accuracy. Eventually, playing darts
more successfully and playing with friends should become naturally reinforcing,
and staff should not have to provide praise any longer. This is the ultimate goal
of training a leisure skill.

204 Chapter 11

Forward Chaining

Forward chaining is similar to backward chaining in that you teach one compo-
nent of the chain at a time and then chain the components together, and you
use prompting and fading to teach the behavior associated with the SD at each
step in the chain. The difference between forward chaining and backward chain-
ing is the point at which you begin training. As you just learned, with backward
chaining, you teach the last component first, then you teach the next to last com-
ponent, and so on; that is, you move from the end of the chain to the front.
In forward chaining, you teach the first component, then the second component,
and so on; that is, you move from the front of the chain to the end.

To use forward chaining, you present the first SD, prompt the correct
response, and provide a reinforcer after the response.

SD1 + prompt � R1 � reinforcer

You then fade your prompts until the person is engaging in the first response
without any prompts when the first SD is presented.

To train the second component, you present the first SD and the learner
makes the first response. Because the first response creates the second SD, you
then prompt the second response and provide a reinforcer after it occurs.

SD1 � R1
SD2 þ prompt � R2 � reinforcer

You fade the prompts until the learner is making the second response without
any prompts. Now, every time you present the first SD the learner makes the first
two responses in the chain.

When you are ready to train the third response in the chain, you present the
first SD and the learner makes the first two responses. The second response creates
the third SD, so as soon as it occurs you prompt the third response and provide a
reinforcer after the response.

SD1 � R1
SD2 � R2

SD3 þ prompt � R3 � reinforcer
Once again, you fade the prompts until the third response occurs when the

third SD is present, without any prompts. Now, every time you present the first
SD the learner makes the first three responses because these three responses have
been chained together through training.

This process of teaching new components continues until you have taught
the last component in the chain and all the steps in the task analysis have been
chained together in the proper order.

Describe how you would use forward chaining to teach the three-step task analy-
sis of eating with a spoon presented earlier in this chapter.

You start by putting a bowl of food (applesauce) and a spoon on the table in
front of the learner. This is the first SD. Now prompt the first response. Take the
learner’s hand, pick up the spoon, put it in the applesauce, and provide a

Chaining 205

reinforcer (praise and, occasionally, a small bite of food). As you feel the learner
start to engage in the behavior with you, fade the prompt until the learner can do
the behavior without any assistance.

Now add step two. Start by presenting the first SD. As soon as the learner
engages in the first response and the spoon is in the bowl (the second SD), physi-
cally prompt the second response—scooping food on the spoon—and provide a
reinforcer after the response. Fade the prompt until the learner can scoop food
on the spoon without any assistance.

Finally, add step three. Again, start by presenting the first SD. As soon as
the learner makes the first two responses and food is scooped onto the spoon
(the third SD), prompt the learner to raise the spoon and put the food into his or
her mouth (third response). The taste of food will be a natural reinforcer for the
third response. Fade your prompts. Now the learner will make all three responses
and eat applesauce with a spoon without any assistance.

Because you provide a reinforcer after each response in the chain during
training, the outcome of each response (the SD for the next response) becomes a
conditioned reinforcer. This is especially important with forward chaining because
you do not get to the natural reinforcer at the end of the chain until you train the
last component. As with backward chaining, once the learner exhibits all the
behaviors in the chain, you eventually switch from a continuous reinforcement
schedule to an intermittent reinforcement schedule to maintain the behavior.
The ultimate goal is to have the behavior maintained by natural reinforcers.

Similarities between Forward Chaining and Backward Chaining
â–  Both are used to teach a chain of behaviors.

â–  To use both procedures, you first have to conduct a task analysis that breaks down the chain into
stimulus-response components.

â–  Both teach one behavior (one component of the chain) at a time and chain the behaviors together.

â–  Both procedures use prompting and fading to teach each component.

Differences between Forward Chaining and Backward Chaining
â–  Forward chaining teaches the first component first, whereas backward chaining teaches the last

component first.

â–  With backward chaining, because you teach the last component first, the learner completes the chain in
every learning trial and receives the natural reinforcer in every learning trial. In forward chaining, the
learner does not complete the chain in every learning trial; artificial reinforcers are used until the last
component of the chain is taught. The natural reinforcer occurs after the last behavior of the chain.

Total Task Presentation

Both forward and backward chaining procedures break down a chain of behaviors
into individual stimulus-response components, teach one component at a time,
and chain the components together. In total task presentation, by contrast,
the complex chain of behaviors is taught as a single unit. As the name of the
procedure implies, the total task is completed in each learning trial.

206 Chapter 11

In total task presentation procedures, you use prompting to get the learner to
engage in the entire chain of behaviors from start to finish. You use whatever type
of prompting strategy is necessary to get the learner to engage in the entire task. In
many cases, physical prompts are used to guide the learner through the chain of
behaviors. Once the learner successfully completes the task with prompts, you
fade the prompts over learning trials until the learner engages in the task without
any assistance. Of course, you provide a reinforcer every time the learner
completes the task, with or without prompts.

One type of physical prompting and fading often used with the total task
presentation procedure is called graduated guidance (Demchak, 1990; Foxx &
Azrin, 1972; Sulzer-Azaroff & Mayer, 1991). With graduated guidance, you use
hand-over-hand guidance to lead the learner through the task. Over trials, you grad-
ually provide less and less assistance and shadow the learner’s hand as the learner
completes the task. Shadowing means keeping your hand close to the learner’s
hand as the learner engages in the behavior. This allows you to initiate physical
guidance immediately if the learner fails to execute one of the component beha-
viors in the chain. Shadowing prevents errors and should be done a number of
times as the learner exhibits the behavioral chain without assistance. As an exam-
ple, consider the use of total task presentation with graduated guidance to teach
a child, Alex, to eat with a spoon. Earlier in this chapter, forward chaining was
illustrated with this same behavior.

To use total task presentation with graduated guidance, you begin by present-
ing the first SD. You put the bowl of applesauce and the spoon on the table in
front of Alex. Next, you use graduated guidance and physically guide Alex through
the entire chain of behaviors. You stand behind Alex, take his hand in yours, put
his fingers around the handle of the spoon, lift his hand with the spoon, put the
spoon in the applesauce, guide his hand to scoop applesauce onto the spoon,
and help him lift his hand with the spoonful of food and put it into his mouth.
You physically guide this chain of behaviors from start to finish. The reinforcer in
each learning trial is the food that Alex eats from the spoon. It is the natural out-
come of the behavior.

After a few trials in which you guide Alex’s hand as he takes a bite of food, he
will start to make some of the behavioral movements himself. As you feel him start
to engage in the behavior, you release his hand and shadow his movements. If
he engages in the correct movements, you continue to shadow his hand. If he
fails to make the correct movement at some point, you start the physical guidance
again. But if you feel him make the correct movement again, you shadow his
hand once again.

For example, as you guide Alex’s hand to pick up the spoon from the table,
you feel him start to put the spoon into the applesauce. You quit guiding
his hand and begin shadowing his hand. Once he has the spoon in the bowl, if
he fails to scoop food onto the spoon, you initiate physical guidance again. Once
the food is on the spoon, if he starts to lift it from the bowl, you stop physically
guiding his hand and start shadowing again. As this process continues, you begin
to shadow more and physically guide less. Eventually, you no longer have to
physically guide the behavior at all. You fade the physical guidance to shadowing,
and you fade the shadowing until you are providing no assistance as Alex takes a
bite of food.

Chaining 207

When to Use Total Task Presentation
â–  Because the total task presentation procedure requires you to guide the learner through the entire

chain of behaviors, it is appropriate for teaching a task that is not too long or too complex. If the task
is too long or difficult, forward or backward chaining procedures may be better because they focus on
one component at a time and chain the components together after they are mastered individually.

■ The learner’s ability level must be considered. Backward or forward chaining may be more
appropriate for learners with limited abilities.

■ Finally, the teacher’s ability level must be considered. Although training is also needed to use forward
chaining and backward chaining successfully, the total task presentation procedure may be the most
difficult to implement. This is because it often involves the use of graduated guidance, a procedure in
which the teacher must alternately guide or shadow the learner with precise timing through the entire
chain of behaviors. Done incorrectly, graduated guidance may amount to forcing the learner through
the behavior without actually teaching the learner to engage in the behavior independently.

To provide graduated guidance correctly, you have to follow Alex’s
movements carefully and respond with more or less guidance as necessary. If you
provide physical guidance too long and do not fade it to shadowing, Alex may
become dependent on the physical prompting and may not learn to engage in
the behaviors himself. That is, if you are going to do it for him, he will not learn
to do it for himself. The goal of any prompting procedure is to fade the prompts
once they are no longer needed. However, you want to fade the physical guidance
to shadowing only as you feel the learner make correct movements. And you want
to initiate the physical guidance again immediately when the learner stops making
the correct behavioral movements. It is important to praise the learner when you
stop physically guiding and begin shadowing. In this way, you will be providing
a reinforcer when the learner engages in the behavior without prompts; thus,
you differentially reinforce independent movements as opposed to prompted
movements. This strengthens the correct behavior and allows you to fade the
physical prompts more quickly.

In some cases, prompting strategies other than graduated guidance may be
used in total task presentation. For example, Horner and Keilitz (1975) used the
total task presentation method to teach toothbrushing to children and adolescents
with intellectual disabilities. They developed a 15-step task analysis of toothbrush-
ing and used 3 types of prompts to teach the behaviors in the task analysis: physi-
cal guidance plus verbal instruction, demonstration plus verbal instruction, and
verbal instruction alone. In every learning trial, the researchers prompted every
step in the task analysis. They used the more intrusive prompts only as needed
and faded the prompts until they were providing no help. Figure 11-2 shows the
graph for the eight subjects in Horner and Keilitz’s study.

Similarities between Forward and Backward Chaining and Total
Task Presentation
â–  They are all used to teach complex tasks or chains of behavior

â–  A task analysis must be completed before training with all three procedures.

â–  Prompting and fading are used in all three procedures.

208 Chapter 11

N
um

be
r o

f T
oo

th
br

us
hi

ng
S

te
ps

C
om

pl
et

ed
C

or
re

ct
ly

0
0 10 3020

Sessions

•
•••

•
•

•• •
•

•
•

•
•

•
•

•

•
••

• ••
•

•
•

•
•

•
••

•
•15

5

10

40

0
0 10 3020

•
•

•
•

•
•

•
• •

••
•

• ••
•

• ••

•

•
•

•
•

•
••• •

•
•

•

•
•

15

5

10

40

0
0 10 3020

•
••

•

•
••

•
•

•

•

•

• ••
•

•
•

••
• •
•• • •••

•

•
•

•
••

•
15

5

10

40

0
0 105 15 25 35

5 15 25 35

5 15 25 35

5 15 25 35 105 15 25

5 15 25

5 15 25

5 15 253020

•

•
•

•

•
•

••
•

•
•

•
•

•
•

•
• •
•

•
•

•••
• ••

•
•

••
• • •

•• •

15

5

10

40

N
um

be
r o

f T
oo

th
br

us
hi

ng
S

te
ps

C
om

pl
et

ed
C

or
re

ct
ly

0
0 3020

Sessions

•
•

•
•

• •
•

•
•

•••
•

•
•

•
•

•
•

•

•
•

•
•

•••15

5

10

0
0 10 3020

•
•

••
•

•
••

•
••• •

•••
•

•
••

• ••
•

15

5

10

0
0 10 3020

•

•

••
•

••
•

• •

•

•

•
••

• •

•
••

• •
•

•
• •
••

15

5

10

0
0 10 3020

•
••

•

•
••

•

• ••• • ••• • ••• •
15

5

10

Baseline Training Baseline Training

Tom

Michael

Larry

Russell

Charles

Jessie

Coleen

Joyce

FIGURE 11-2 This graph shows the number of toothbrushing steps completed correctly by eight children and
adolescents with intellectual disabilities. Total task presentation was used to teach this task. This
graph shows a multiple-baseline-across-subjects research design. The implementation of treatment
was staggered over time for each subject; the number of toothbrushing steps completed correctly for
each subject increased only after the treatment was implemented. (From Horner, R. H., & Keilitz, I.
[1978]. Training mentally retarded adolescents to brush their teeth. Journal of Applied Behavior
Analysis, 8, 301–309. Copyright 1978 Society for the Experimental Analysis of Behavior. Reprinted
by permission of the Society for the Experimental Analysis of Behavior.)

Chaining 209

Difference between Forward and Backward Chaining and Total
Task Presentation
â–  In total task presentation, the learner is prompted through the entire task in each learning trial. In
the two chaining procedures, the trainer teaches one component of the chain at a time and then
chains the components together.

Other Strategies for Teaching
Behavioral Chains

Teaching complex tasks by forward chaining, backward chaining, or total task
presentation entails substantial trainer time in performing the prompting and
fading procedures with the learner. Other strategies to teach complex tasks take
less trainer time and involvement. These strategies—written task analysis, picture
prompts, video modeling and self-instructions—involve the independent use of
prompts to guide appropriate completion of the chain of behaviors.

Written Task Analysis
For people who have the ability to read, written task analysis can be used to
guide appropriate performance of a chain of behaviors. In this strategy, the trainer
presents the learner with a list of the component behaviors in their proper
sequence, and the learner uses this list to perform the task correctly. For example,
when you buy a stereo system, the assembly instructions guide you through
the task of assembling it. The list of instructions is a written task analysis. A written
task analysis is effective only if the learner can read the instructions, understand
the instructions, and execute the behavior listed in the instructions. To be most
effective, the written task analysis must clearly and specifically list every component
behavior in the chain.

Cuvo and his colleagues (Cuvo, Davis, O’Reilly, Mooney, & Crowley, 1992)
used written task analyses (also called textual prompts) to teach young adults with
mild mental disabilities and learning disabilities to clean household appliances
such as a stove and refrigerator. They gave the learners a detailed list of all the
steps involved in the tasks (a written task analysis). The learners used the detailed
list to guide their behavior of cleaning the appliances. When they were finished,
they received praise for correct performance or corrective feedback (further
instructions) on how to improve if they made errors. The researchers found that
all the learners performed the tasks correctly with the use of the written task
analyses and reinforcement for correct performance.

Picture Prompts
Another strategy used to guide appropriate performance of a chain of behaviors is
the use of picture prompts. With picture prompts, you take pictures of the out-
come of each behavior or of someone engaging in each behavior in the task.
The pictures are then used to prompt the learner to engage in the behaviors
in the proper sequence. To be effective, the learner has to look at the pictures in

210 Chapter 11

the proper sequence, and each picture must have stimulus control over the
behavior that is depicted. Consider the following example.

Saul, a worker with intellectual disability, has a job in a company that does pro-
motional mailings. Saul’s job is to put brochures into envelopes for mailing. The
company mails 20 different brochures, and Saul has to put 3 to 6 of them in a
large envelope, depending on the job for that day. The job trainer has pictures of
all 20 brochures. At the beginning of each work day, the trainer gets the pictures
of the brochures that Saul has to put in envelopes for that day. The trainer tacks
the pictures onto a board at Saul’s workstation; Saul looks at the pictures to help
him put the correct brochures in the envelope (Figure 11-3). The picture prompts
have stimulus control over the behavior of choosing the correct brochures. After
the job trainer sets up the picture prompts, the trainer does not have to spend any
further time using prompting and fading to teach the task to Saul.

Wacker and his colleagues (Wacker, Berg, Berrie, & Swatta, 1985) used
picture prompts to teach adolescents with severe disabilities to complete complex
vocational and daily living tasks such as folding laundry or assembling industrial
parts. The researchers put pictures of each step in the tasks into notebooks and
taught the adolescents to turn the pages of the notebook to see the picture
prompts. All three adolescents in the study learned to use the picture prompts in
the notebooks to guide their behavior. Once they learned to use the picture
prompts, they did not need any further prompting to complete the tasks.

Video Modeling
Another strategy for teaching a learner to engage in a chain of behaviors is video
modeling or video prompting. In this teaching strategy, the learner watches a
video of someone engaging in the chain of behaviors immediately before engaging

FIGURE 11-3 Saul is completing a work task with the assistance of picture prompts on the bulletin board in front
of him. Each picture acts as a prompt for completion of each component of the work task.

Chaining 211

in the same task. By watching the task being completed on video, the learner is
then able to complete the chain of behaviors. Video modeling has been used to
teach learners with intellectual disabilities to engage in a variety of skills including
laundry skills (Horn et al. 2008), meal preparation skills (Rehfeldt, Dahman,
Young, Cherry, & Davis, 2003), dishwashing skills (Sigafoos et al., 2007) and
microwave use (Sigafoos et al., 2005).

The video modeling procedure may be conducted in two different ways. In
one, the learner watches the entire video just before attempting to complete the
task (e.g., Rehfeldt et al.) and in the other, the learner watches one step of the task
on video, completes that task, then watches the next step on video and completes
that step, and then proceeds in this manner until the task is completed (e.g., Horn
et al.). In the study by Horn et al., three individuals with intellectual disabilities
completed a 10-step laundry task after watching a video of the task. However, the
task had to be broken down on video in a different number of steps for each partici-
pant. For example, one individual could only complete the task by watching each
step on video and then completing that step before moving to the next step.
Another individual could complete the first 5 steps of the chain after watching the
first 5 steps on video and then complete the second 5 steps after watching those
steps on video. This study showed that learners may need the video of the task bro-
ken down into a different number of steps before they can complete the chain.

Self-Instructions
Learners can also be guided through a complex task by means of self-generated
verbal prompts (also called self-instructions). In this procedure, you teach the
learners how to give themselves verbal prompts or instructions to engage in the
correct sequence of behaviors in the chain. To use this procedure, the learners
must be able to remember the self-instructions, say them at the appropriate time,
and correctly follow the self-instructions. (The self-instructions must have stimulus
control over the behavior.) The learner first learns to recite the self-instructions
out loud as a prompt for the correct behavior. After the learner has mastered the
self-instructions, he or she may then begin to recite the self-instructions covertly.
You might think that a person capable of learning the self-instructions could also
learn the behaviors in the chain, and thus the self-instructions would not be nec-
essary. Although this may be true for many people, some learners who have diffi-
culty completing a complex task may benefit from self-instructions. In addition,
because self-instructions can be recited quickly and remembered easily in many
cases, they are useful for prompting behavior in a variety of situations.

Consider the following everyday examples of self-instructions. Each time you
go to your locker and recite the combination on the lock as you open it, you are
using self-instructions. When you recite the seven digits of a phone number as you
dial the numbers, you are using self-instructions. When you talk yourself through
the steps in a recipe (“I need to add 2 cups of flour, 1 cup of oats, 1 cup of raisins,
and a teaspoon of baking powder”), you are using self-instructions to prompt each
behavior in a behavioral chain.

A number of studies have demonstrated that learners can use self-instructions to
guide themselves through complex vocational or educational tasks. For example,
Salend, Ellis, and Reynolds (1989) taught adults with severe intellectual disabilities

212 Chapter 11

to recite self-instructions to prompt the correct sequence of behaviors in a vocational
task (packaging plastic combs). They said four simple self-instructions: “Comb up,
comb down, comb in bag, comb in box.” As they said each self-instruction, they
completed the task associated with that instruction. The use of the self-instructions
led to correct performance of the task. Whitman, Spence, and Maxwell (1987)
taught adults with intellectual disabilities to use self-instructions to prompt the
behavior of correctly sorting letters into boxes. Albion and Salzburg (1982) taught
students with mental disabilities to use self-instructions to complete math problems
correctly. In each case, the self-instructions prompted the correct behaviors in the
chain to occur in the correct sequence.

Although written task analyses, picture prompts, video modeling and self-
instructions are prompting strategies that are often used to teach a chain of beha-
viors, they may also be used with single responses as well. The procedures are
described in this chapter to illustrate their use with chains of behaviors.

Chaining Procedures
â–  Backward chaining: Teach the last behavior in the chain first; then teach each previous behavior

in the chain

â–  Forward chaining: Teach the first behavior in the chain first; then teach each subsequent behavior
in the chain

â–  Total task presentation: Prompt the whole stimulus-response chain in each learning trial
â–  Written task analysis: Use written descriptions of each step in the task analysis as prompts.
â–  Picture prompts: Use pictures of each step in the task analysis as prompts.
â–  Video modeling: Use video of the task (or parts of the task) as prompts to complete the task.
â–  Self-instructions: Give yourself verbal prompts to engage in each component behavior in a behav-

ioral chain

FOR FURTHER READING
Varied Applications of Chaining Procedures

Many everyday activities and tasks are composed of behavioral chains, and research has demonstrated the
effectiveness of chaining procedures for teaching a wide range of activities. For example, Thompson, Braam,
and Fuqua (1982) used a forward chaining procedure to teach laundry skills to individuals with intellectual
disabilities. They conducted a task analysis of the behaviors involved in operating a washer and dryer and
showed that three individuals learned to engage in the chain of behaviors composed of 74 individual
responses. In another study, MacDuff, Krantz, and McClannahan (1993) used picture prompts to facilitate
the completion of complex leisure and homework activities by four children with autism. The children learned
to use a three-ring binder containing pictures of the activities to be completed. Each child looked at the
sequence of pictures and completed the activities depicted in the pictures. Although the children were capable
of completing the activities before the use of picture prompts, they did not consistently do the activities until
trained to use the picture prompts. A study by Vintere and colleagues (Vintere, Hemmes, Brown, & Poulson,
2004) demonstrated the effectiveness of a total task presentation procedure for teaching complex dance steps
to preschool children. In this study, the authors used instructions and modeling to prompt the children to
engage in the chain of behaviors and provided praise for correct performance. Some children received self-
instruction in addition to instructions and modeling. The authors showed that both procedures were effective,
but that the addition of self-instructions resulted in the children learning the dance step more rapidly.

Chaining 213

How to Use Chaining Procedures

If your goal is to teach a person a complex task, you may use one of the proce-
dures described in this chapter. All of the procedures described here are consid-
ered chaining procedures because they are used to teach a chain of behaviors.
Thus, in the present context, chaining procedure is an inclusive term that refers
to backward and forward chaining, total task presentation, written task analysis,
picture prompts, video modeling, and self-instructions. The following steps are
important for the effective use of chaining procedures (see also Cooper et al.,
1987, 2007; Martin & Pear, 1992; Sulzer-Azaroff & Mayer, 1991).

1. Determine whether a chaining procedure is appropriate. Does the problem
call for behavioral acquisition, or is it related to noncompliance? If the
person is not completing a complex task because he or she is not capable,
a chaining procedure is appropriate. Conversely, if the person is capable of
completing the task but is refusing to engage in it, procedures for treating
noncompliance are warranted.

2. Develop a task analysis. The task analysis breaks down the chain of
behaviors into individual stimulus-response components.

3. Get a baseline assessment of the learner’s ability. Cooper and colleagues
(1987) describe two methods for assessing the mastery level of the learner. In
the single-opportunity method, you present the learner with the opportunity
to complete the task and record which components the learner completes
without assistance in the correct sequence. That is, you present the first SD

and assess the learner’s responses. The first error by the learner in single-
opportunity assessment will typically result in errors on all subsequent steps in
the task analysis or in the learner’s inability to complete any further steps.
In the multiple-opportunity method, you assess the learner’s ability to com-
plete each individual component in the chain (e.g., Horn et al., 2008). You
present the first SD and wait for the learner to respond. If the learner does not
respond correctly, you present the second SD and assess the learner’s response.
If there is no correct response, you present the third SD, and so on, until the
learner has had the opportunity to respond to every SD in the chain.

4. Choose the chaining method you will use. For learners with the most limited
abilities, forward or backward chaining methods are most appropriate. If the
task is less complex or if the learner is more capable, total task presentation
may be more appropriate. Other procedures such as written task analysis,
picture prompts, video modeling, or self-instructions may be appropriate,
depending on the capabilities of the learner or the complexity of the task.

5. Implement the chaining procedure. Whichever procedure you use, the ulti-
mate goal is to get the learner to engage in the correct sequence of behaviors
without any assistance. Therefore, the appropriate use of prompting and
fading is important in all the chaining procedures. Continue to collect data
on the learner’s performance as you implement the chaining procedure.

6. Continue reinforcement after the task has been learned. If you continue to pro-
vide reinforcement, at least intermittently, after the learner is able to complete
the task without assistance, the learner will maintain the behavior over time.

214 Chapter 11

CHAPTER SUMMARY

1. A behavioral chain, also called a stimulus-
response chain, is a behavior composed of two
or more stimulus-response components.

2. A task analysis identifies the stimulus and
response in each component of the chain. It is
important to conduct a task analysis so that all
components of the chain (SDs and responses)
are identified clearly.

3. Chaining procedures are used to teach a person
to engage in a behavioral chain. These proce-
dures involve prompting and fading to teach
each component of the chain. In backward
chaining, the last stimulus-response component
is taught first. The next to last component is
taught next, and so on, until the whole chain is
learned. In forward chaining, the first stimulus-
response component is taught first. The second

component is taught second, and so on, until the
whole chain is learned.

4. In total task presentation, the entire chain of
behaviors is prompted in every learning trial.
Often, graduated guidance is used with total task
presentation.

5. In the written task analysis procedure, the learner
uses textual prompts for each component in the
chain. In the picture prompt procedure, the
learner uses pictures to prompt each component
in the behavioral chain. In the video modeling
procedure, the learner watches video of the task
being completed to prompt the completion of the
steps in the task. With self-instructions, the learner
recites self-instructions (verbal prompts) to prompt
each component in the chain.

KEY TERMS

backward chaining, 202
behavioral chain, 197
chaining procedures, 202
forward chaining, 205

graduated guidance, 207
picture prompts, 210
self-instructions, 212
stimulus-response chain, 198

task analysis, 199
total task presentation, 206
video modeling, 211
written task analysis, 210

PRACTICE TEST

1. What is a stimulus-response chain? Provide two
examples of stimulus-response chains that are
not in this chapter. (pp. 197–198)

2. Identify each stimulus and response component
in your two examples from Question 1. (pp.
197–198)

3. What is a task analysis? Why is it important to
conduct a task analysis? (p. 199)

4. Provide a task analysis of the behavior of pouring
water from a pitcher into a glass. Assume the
pitcher of water and the glass are already on
the table. (pp. 199–202)

5. Describe backward chaining. (pp. 202–204)
6. Describe the use of backward chaining to teach

the task identified in Question 4. (pp. 202–204)
7. Describe forward chaining. (pp. 205–206)

8. Describe how you would use forward chaining
to teach the task identified in Question (pp.
205–206)

9. How are backward chaining and forward chain-
ing similar? How are they different? (p. 206)

10. Describe the total task presentation procedure.
(pp. 206–208)

11. Describe graduated guidance. (p. 207)
12. Describe how you would use the total task

presentation procedure to teach the task
described in Question 4. (pp. 206–208)

13. How does the total task presentation procedure
differ from backward and forward chaining?
How are they similar? (pp. 208, 210)

14. Describe how you would use a written task anal-
ysis to get a person to engage in a complex task.

Chaining 215

What is another name for a written task analysis?
(p. 210)

15. Describe the use of picture prompts. Describe
the use of video modeling. (pp. 210–212)

16. Describe the use of self-instructions. What is
another name for self-instructions? (pp. 212,
213)

17. When is it appropriate to use a chaining proce-
dure? When is it not appropriate? (p. 214)

18. Briefly describe the guidelines for using a
chaining procedure to teach a complex task.
(p. 214)

APPLICATIONS

1. You have been hired by an agency that provides
rehabilitation services for people who have sus-
tained brain damage from head injuries. These
people often have to learn basic skills all over
again. One skill that you have to teach is bed-
making. Your first step is to develop a task anal-
ysis of bed-making. Provide the task analysis for
bed-making. Be sure to include all the stimulus-
response components.

2. Once you have developed the task analysis for
the complex task of bed-making, you must
choose a chaining procedure and implement

the procedure. You decide to use forward chain-
ing. Describe the use of forward chaining to
teach the bed-making task.

3. One of the people with brain damage has a seri-
ous memory impairment. A day after he learns
the task, he cannot remember the bed-making
behaviors. You decide that you will use either
picture prompts or textual prompts (written task
analysis) to help him make his bed each day.
Describe how you would use picture prompts
and how you would use textual prompts with
this person.

MISAPPLICATIONS

1. Your niece has just been enrolled in a preschool
program by her parents. Before she starts, you
want to teach her to recite the alphabet. Because
reciting the alphabet is a chain of behaviors, you
decide that you will use graduated guidance to
teach her. What is wrong with the use of gradu-
ated guidance in this situation? What would be
a better procedure for teaching her to recite the
alphabet?

2. Toby, a young man with severe intellectual dis-
ability, recently started a job in which he assem-
bles parts for bicycle brakes. The task has seven
steps. Staff members used picture prompts to
help Toby learn the task, and they used tokens
to reinforce the behavior. At the end of every
month, Toby gets a paycheck based on the num-

ber of parts he assembles. Once Toby learned
the job, staff removed the picture prompts and
stopped using the tokens. Now they just let Toby
do his work and expect that his monthly pay-
check will maintain the behavior. What is the
problem with this strategy? What would be a
better strategy?

3. Waylon, a college student, is home for the sum-
mer and has just started a job at a retail store in
the mall. He works the evening shift and has to
close the store and lock up for the evening.
There is a list of 20 steps involved in closing
and locking up. The manager decides he will
use forward chaining to teach the task to
Waylon. What is the problem with this strategy?
What would be a better strategy?

216 Chapter 11

Chapter Twelve

Behavioral Skills Training Procedures

You have learned about prompting and fading procedures that may be used toteach a person to engage in the correct behavior at the right time (to establish
stimulus control over the behavior). You have also learned about chaining procedures,
in which prompting and fading are used to teach a person a complex task. In this
chapter, you will learn other procedures for teaching skills. Four behavioral skills
training (BST) procedures—modeling, instructions, rehearsal, and feedback—
generally are used together in training sessions to help a person acquire useful skills

(such as social skills or job-related skills). BST procedures are typically
used to teach skills that can be simulated in a role-play context.

Examples of Behavioral Skills
Training Procedures

Teaching Marcia to Say “No” to the Professors
Marcia is a secretary at a university. She believes that faculty members
in her department make unreasonable demands on her, but she has
not been able to refuse these unreasonable requests (such as working

through her lunch hour and running personal errands). She is seeing a psychologist,
Dr. Mills, who is using BST procedures to help her develop assertiveness skills. In
the psychologist’s office, they role-play the difficult situations that Marcia faces at
work. Dr. Mills uses the role-plays to assess Marcia’s assertiveness skills and to teach
her how to act more assertively. First, Dr. Mills creates a situation at work in which
Marcia role-plays herself and he role-plays a coworker. In that role, he makes an
unreasonable request, such as “Marcia, I have a meeting this afternoon. I need you
to go pick up my dry cleaning on your lunch hour.” He then assesses what she says
and how she says it (her verbal and nonverbal behavior) in response to this request.
Next, Dr. Mills provides instructions and modeling; that is, he describes how to
respond more assertively in this situation and demonstrates the assertive behavior
for Marcia in another role-play. This time Marcia role-plays the coworker making
the unreasonable request, and Dr. Mills plays Marcia responding assertively. In the
role-play, Dr. Mills says, “I’m sorry, but I can’t do your personal errands for you.”

â–  What are the four components of a
behavioral skills training procedure?

â–  When is the appropriate time to use
behavioral skills training procedures?

â–  How do you use behavioral skills
training procedures in groups?

â–  How is the concept of the three-term
contingency related to behavioral skills
training procedures?

217

After observing Dr. Mills model this assertive behavior, Marcia gets an opportunity
to practice it (rehearsal): They switch roles again, and Marcia makes the same asser-
tive response in the role-play. Dr. Mills then gives her feedback on her performance.
He praises her for the aspects of the behavior that she performed well, and he gives
her suggestions on how to improve. After getting the feedback, Marcia practices the
behavior again in another role-play. Again Dr. Mills praises Marcia for her perfor-
mance and makes any necessary suggestions for improvement. Once Marcia has
learned this assertive behavior well, they will role-play other situations that arise at
work. Marcia will learn a variety of assertiveness skills through this process of instruc-
tions, modeling, rehearsal, and feedback.

Teaching Children to Protect Themselves from Abduction
Consider another example. Cheryl Poche used modeling, instructions, rehearsal,
and feedback to teach abduction prevention skills to preschool children (Poche,
Brouwer, & Swearingen, 1981). She taught the children how to respond to adults
who tried to lure the children into leaving with them. Poche set up realistic role-
plays in which an adult walked up to the child on the playground and asked the
child to leave with him. The adult would say something like, “Hi, I have a toy in
my car I think you’d like. Come with me, and I’ll get it for you.” The skills that
the children learned were to say, “No, I have to ask my teacher,” and to run back
into the school. First, Poche used the role-plays to assess the children’s skills before
training. Next, she implemented the BST procedure. The child watched as two adult
trainers acted out a scene in which one trainer, playing the suspect, walked up and
asked the other trainer, playing the child, to leave with him. The trainer playing the
child then modeled the correct response to this lure. After watching the model, the
child practiced the abduction prevention skill in another role-play. A trainer
approached the child and presented the abduction lure. In response, the child said,
“No, I have to ask my teacher,” and ran back to the school (Figure 12-1). The trainer
praised the child for correct performance and, if the response was only partly correct,
the trainer provided instructions and further modeling.

The child rehearsed the behavior again in role-plays until the behavior was
correct. Then the child received training with different types of abduction lures
until the child could make the correct response in a variety of situations. The
results of this study are illustrated in Figure 12-2.

Components of the Behavioral Skills
Training Procedure

As you can see from these examples, four procedures are commonly used together
to teach skills. Let’s examine these procedures in more detail.

Modeling
With modeling, the correct behavior is demonstrated for the learner. The learner
observes the model’s behavior and then imitates the model. For modeling to be effec-
tive, the learner must have an imitative repertoire; that is, the learner has to be able to
pay attention to the model and perform the behavior that the model just demonstrated.

218 Chapter 12

Most people have imitative repertoires because imitating the behavior of
others has already been reinforced in a variety of situations (Baer, Peterson, &
Sherman, 1967). Reinforcement for imitation typically starts early in a child’s life.
Over the course of early development, a child’s behavior of imitating models
(provided by parents, teachers, siblings, and peers) is reinforced many times in
the presence of a wide variety of behaviors modeled by a variety of people. As a
result, a model’s behavior becomes an SD for imitation, and imitation becomes a
generalized response class, which means that imitation is likely to occur in the
future when a behavior is modeled for the learner (Baer & Sherman, 1964;
Bijou, 1976; Steinman, 1970).

Modeling may be live or it may be symbolic. In live modeling, another per-
son demonstrates the appropriate behavior in the appropriate situation. With sym-
bolic modeling, the correct behavior is demonstrated on videotape, audiotape, or
possibly in a cartoon or a movie. For example, in another study by Poche, Yoder,
and Miltenberger (1988), grade school children viewed a videotape in which
abduction prevention skills were demonstrated by child actors. The videotape
showed an adult approach a child and present an abduction lure. The child then
engaged in the correct behavior in response to the abduction lure. The model’s
behavior in the videotape was the same as the behavior of the live model in
Poche’s earlier study. In this study, however, a whole class of children viewed the
videotaped model at one time. The videotape also included instructions about
the correct behavior. After the children viewed the videotape, they rehearsed the
correct behavior and received praise or further instruction if they needed it.
Another group of children viewed the videotape but did not rehearse the behavior.

FIGURE 12-1 After receiving the abduction solicitation from the adult, the child says, “No, I have to ask my
teacher,” and runs back to the school. The trainer praises the child for exhibiting the skill correctly.

Behavioral Skills Training Procedures 219

Le
ve

l o
f A

pp
ro

pr
ia

te
ne

ss
o

f S
el

f-
Pr

ot
ec

tiv
e

Re
sp

on
se

s

0

5

1

2

4

6

3

0 5 10 15 20 25
Days

Baseline Training Generality Follow-Up

0

5

1

2

4

6

3

0

5

1

2

4

6

3

•

•

•

•

•
•

•
Patty

Stan

John

School Grounds Community

Type of Lure
Simple
Authority
Incentive

•

FIGURE 12-2 This graph shows the level of self-protection skills before and after a behavioral skills training
procedure was implemented with three preschool children. The self-protection skills were rated
on a scale from 0 to 6. A score of 6 means that the child said, “No, I have to ask my teacher,”
and ran back to school when an abduction lure was presented. A score of 0 means that the
child agreed to leave with the adult who presented the abduction lure. Sometimes the child
was assessed on the playground and sometimes in the community. Three types of lures were
used: With a simple lure, the adult simply asked the child to leave with him; with an authority
lure, the adult said that the child’s teacher said it was okay for the child to leave with him;
with the incentive lure, the adult offered the child something like a toy if the child would leave
with him. This graph shows a multiple-baseline-across subjects design in which three children
received training at different times. (From Poche, C, Brouwer, R., & Swearengin, M. [1981].
Teaching self-protection to young children. Journal of Applied Behavior Analysis, 14, 169–176.
Copyright © 1981 Society for the Experimental Analysis of Behavior. Reprinted by permission of
the Society for the Experimental Analysis of Behavior.)

220 Chapter 12

The researchers found that the children who received the modeling, instructions,
rehearsal, and feedback learned the abduction prevention skills better than
the children who got instructions and modeling from the videotape without
the chance for rehearsal and feedback.

A number of factors influence the effectiveness of modeling (Bandura, 1977).

â–  When the model exhibits the correct behavior, it should result in a successful
outcome (a reinforcer) for the model.

â–  The model should resemble the people observing the model or should have
high status. For example, the models in Poche’s videotape were children of the
same age as those watching the tape. Often, teachers model correct behavior for
children. Because teachers have high status, the children are likely to learn from
the model. In television commercials, typically sports stars and other celebrities
(people of very high status) are shown using the product. The hope is that people
will imitate the model and buy the product.

■ The complexity of the model’s behavior should be appropriate to the devel-
opmental or ability level of the learner. If the model’s behavior is too complex, the
learner may not be able to learn from it. However, if the model’s behavior is too
simple, the learner may not pay attention.

â–  The learner has to pay attention to the model to learn the behavior being
modeled. Often, the teacher will draw the learner’s attention to important aspects
of the model’s behavior. When modeling assertiveness skills, Dr. Mills focused
Marcia’s attention by saying, “Now watch how I make eye contact and use a firm
tone of voice.” In Poche’s videotape, the narrator told the children what behaviors
to look for each time a model was about to be presented.

â–  The modeled behavior must occur in the proper context (in response to
the relevant SD). The behavior should be modeled in the real situation or in the
context of a role-play of the real situation. For example, the children saw the
abduction skills modeled in response to abduction lures from an adult, that is, in
the situation in which they would be needed. Marcia watched Dr. Mills model
assertive behavior in the context of role-play of difficult interactions Marcia
faced at work.

â–  The modeled behavior should be repeated as often as necessary for the
learner to imitate it correctly.

â–  The behavior should be modeled in a variety of ways and in a variety of
situations to enhance generalization.

â–  The learner should have an opportunity to rehearse (imitate) the behavior
as soon as possible after observing the model. Correct imitation of the modeled
behavior should be reinforced immediately.

Instructions
Instructions describe the appropriate behavior for the learner. To be most effec-
tive, instructions should be specific. They should describe exactly the behaviors
that are expected from the learner. For a chain of behaviors, the instructions
should specify each component in the chain in proper sequence. Instructions
should also specify the appropriate circumstances in which the learner is expected
to engage in the behavior. For example, when teaching abduction prevention

Behavioral Skills Training Procedures 221

skills to young children, the teacher might give this instruction: “Whenever any
adult asks you to leave with him or when an adult asks you to go somewhere
with him, you should say, ‘No, I have to ask my teacher,’ and run back into the
school. You should run in and tell me right away and I’ll be very proud of you.”
This instruction specifies the antecedent situation and the correct behavior.
It also specifies the consequence (teacher approval). The following factors may
influence the effectiveness of instructions.

â–  The instructions should be presented at a level that the learner can under-
stand. If they are too complex, the learner may not grasp the behavior. If they are
too simple, the learner may be indignant or offended.

â–  The instructions should be delivered by someone who has credibility with
the learner (such as a parent, teacher, employer, or psychologist).

â–  The learner should have the opportunity to rehearse the behavior as soon as
possible after receiving the instructions.

â–  Instructions should be paired with modeling whenever observing the
behavior will enhance the potential for learning the behavior.

â–  The instructions should be given only when the learner is paying attention.
â–  The learner should repeat the instructions so that the teacher can be

certain the learner heard the instructions correctly. Repeating the instructions
during training also increases the likelihood that the learner will be able to repeat
the instructions later to self-prompt the appropriate behavior.

Rehearsal
Rehearsal is the opportunity for the learner to practice the behavior after receiving
instructions or watching a model demonstrate the behavior. Rehearsal is an impor-
tant part of the BST procedure because (a) the teacher cannot be sure that the
learner has learned the behavior until the teacher sees the learner engage in the
correct behavior, (b) it provides an opportunity to reinforce the behavior, and (c)
it provides an opportunity to assess and correct errors that may be present in the
performance of the behavior. The following factors may influence the effective-
ness of rehearsal as part of the BST procedure.

â–  The behavior should be rehearsed in the proper context, either in the situa-
tion to which it is appropriate or in a role-play that simulates that situation.
Rehearsing the behavior in the proper context facilitates generalization when skills
training is complete.

â–  Rehearsals should be programmed for success. Learners should practice
easy behaviors (or easy situations in which the behavior should occur) first so that
they are successful. After success with easy behaviors, the learners can practice
more difficult or complex behaviors. In this way, engaging in the rehearsal is
reinforcing, and the learners continue to participate.

â–  Rehearsal of the correct behavior should always be followed immediately by
reinforcement.

â–  Rehearsals that are partly correct or are incorrect should be followed by
corrective feedback.

â–  The behavior should be rehearsed until it is demonstrated correctly at least
a few times.

222 Chapter 12

Feedback
Following the learner’s rehearsal of the behavior, the trainer should provide
immediate feedback. Feedback involves praise or other reinforcers for correct
performance. When necessary, it may also involve correction of errors or further
instruction in how to improve performance. Feedback often amounts to differen-
tial reinforcement of some aspects of the behavior with correction of other aspects.
In BST procedures, feedback is specifically defined as the delivery of praise
for correct performance and further instruction after incorrect performance.
A number of factors may influence the effectiveness of feedback.

â–  Feedback should be given immediately after the behavior.
â–  Feedback should always involve praise (or other reinforcers) for some

aspect of the behavior. If the behavior was not correct, the trainer should praise
the learner at least for trying. The point is to make the rehearsal a reinforcing
experience for the learner.

â–  Praise should be descriptive. Describe what the learner said and did that
was good (correct). Focus on all aspects of the behavior, verbal and nonverbal
(i.e., what the learner said and did and how the learner said and did it).

â–  When providing corrective feedback, do not be negative. Do not describe the
learner’s performance as bad or wrong. Rather, provide instructions that identify what
the learner could do better or how the learner could improve the performance.

â–  Always praise some aspect of the performance before providing corrective
feedback.

â–  Provide corrective feedback on one aspect of the performance at a time.
If the learner did a number of things incorrectly, focus first on one of them so
that the learner does not feel overwhelmed or discouraged. Build the correct
performance in steps so that the learner is more and more successful in each
subsequent rehearsal.

Enhancing Generalization after Behavioral
Skills Training

The goal of BST procedures is for the learner to acquire new skills and to use
these skills in the appropriate circumstances outside the training sessions. Several
strategies can be used to promote generalization of the skills to the appropriate
circumstances after BST.

First, training should involve a variety of role-plays that simulate the actual
situations the learner is likely to encounter in real life. The closer the training
scenarios (role plays) are to the real-life situations, the more likely the skills are to
generalize to the real situations (Miltenberger, Roberts, et al., 1999).

Second, incorporate real-life situations into training. The learner may
rehearse the skills in role-plays with real peers or in real situations (e.g., at school,
on the playground). For example, Olsen-Woods, Miltenberger, and Forman
(1998) taught abduction prevention skills to children and conducted some role-
plays out on the playground of their school as a real-life situation in which an
abduction attempt may take place.

Behavioral Skills Training Procedures 223

Third, provide assignments for the learner to practice the skill being learned
outside the BST session, in a real-life situation. After practicing the skill outside
the training session, the learner can discuss the experience in the next BST
session and receive feedback on his or her performance. In some cases, practice
of the skills outside of a session can be supervised by a parent or teacher who can
provide immediate feedback.

Fourth, the trainer can arrange for reinforcement of the skills in situations
outside the training sessions. For example, the trainer might talk to a teacher or
parent and have him or her provide reinforcement when the learner exhibits the
correct skill at home or school.

In Situ Assessment

Behavioral skills training often occurs in a situation that is different from where
the skills need to be used. For example, abduction prevention training might
occur in a classroom but the skills need to be used out in public when the child
is alone in a situation where an abductor might present a lure. Therefore, it is
important to assess the skills taught with BST in the setting where the skills need
to occur. In addition, it is important to assess the skills without the individual’s
knowledge that an assessment is taking place. When an assessment of skills occurs
in the natural environment where the skills are needed and the individual is not
aware that an assessment is taking place, it is called an in situ assessment.
Conducting an in situ assessment is important for the accurate assessment of
whether the individual will use the skills when they are needed. Research has
shown that if the individual knows that an assessment is taking place he or she
might be more likely to use the skills than if he or she does not know that an
assessment is taking place (Gatheridge et al., 2004; Himle, Miltenberger, Gather-
idge, & Flessner, 2004; Lumley, Miltenberger, Long, Rapp, & Roberts, 1998).

For example, in the study by Gatheridge et al. (2004), 6 and 7-year-olds were
taught safety skills to use if they found a gun and no adult was around: Do not
touch the gun, run away from it, and go tell a parent. After training, when the
children were asked what to do when they found a gun, they provided the correct
response. When they were asked to show the researcher what to do when they
found a gun, they demonstrated the correct behavior for the researcher. However,
when they found a gun (a disabled gun provided by the police dept for use in the
research) without knowing that anyone was watching (in situ assessment), they
failed to exhibit the correct response. The children had learned the skills but did
not use the skills unless the researcher was present; the skills failed to generalize
because they were under the stimulus control of the presence of the researcher.

Why would the skills be under the stimulus control of the researcher (Why is the
researcher an SD for the use of the skills)?

The researcher was an SD for the use of the skills because the skills were
reinforced only when the researcher was present during training. To get the skills
to generalize, often it is necessary to reinforce the skills in the natural environ-
ment when the researcher is not present. This procedure, called in situ training,
is described below.

224 Chapter 12

In Situ Training

Recent research evaluating BST for teaching safety skills to children and indivi-
duals with intellectual disabilities has shown that a procedure called in situ
training is sometimes needed for promoting generalization after training (e.g.,
Egemo Helm et al., 2007; Himle, Miltenberger, Flessner, & Gatheridge, 2004;
Miltenberger et al., 2005; Miltenberger, Roberts et al., 1999). With in situ training,
the trainer sets up an assessment in the natural setting without the child’s knowl-
edge that he or she is being assessed (an in situ assessment). If the child does not
perform the skills during this in situ assessment, a trainer enters the situation and
immediately turns the assessment into a training session. The trainer then has the
child rehearse the skills in the assessment situation so that the skills are more
likely to occur if the child is faced with a similar situation in the future.

Consider the example from a 2005 study by Johnson and colleagues evaluat-
ing BST for teaching abduction prevention skills to 4- and 5-year-olds. After a
5-year-old child showed that she could engage in abduction prevention skills dur-
ing BST sessions, Johnson conducted an in situ assessment. During this assess-
ment, a research assistant (who the child did not know) approached the child on
the playground while the child was alone and asked the child if she would like to
go for a walk. When the child did not use the safety skills (she did not say “No,”
and then run away and tell an adult) during the assessment, a trainer walked out-
side at that moment and asked the child, “What are you supposed to do when a
stranger asks you to leave?” After the child answered with the correct response,
the trainer said, “Well, you didn’t do that. We are going to have to practice so
you do it right if this ever happens again.” The trainer then had the child practice
saying “no,” running away, and telling an adult in response to a role-play in the
actual situation where the assessment took place. As a result, the next time the
child was assessed without her knowledge, she engaged in the correct behaviors.
Across a number of studies, researchers have shown that conducting in situ train-
ing in this fashion is effective for children who did not use the skills after BST
(Gatheridge et al., 2004; Himle, Miltenberger, Flessner, & Gatheridge, 2004;
Johnson et al., 2005, 2006; Jostad et al., 2008; Miltenberger et al., 2004, 2005).

Behavioral Skills Training and the
Three-Term Contingency

By combining modeling, instructions, rehearsal, and feedback, the BST procedure
uses all three aspects of the three-term contingency. A three-term contingency—
involving antecedents, the behavior, and consequences of the behavior—should
be used in any teaching situation. The modeling and instructions are antecedent
strategies used to evoke the correct behavior. Because most people have success-
fully followed instructions or imitated models in the past, instructions and model-
ing are effective discriminative stimuli for the correct behavior. Rehearsal involves
executing the behavior that was modeled or described in the instructions. When
the behavior is rehearsed correctly, feedback involves a reinforcing consequence
that strengthens the correct behavior. When the behavior is partly incorrect,

Behavioral Skills Training Procedures 225

corrective feedback is provided in the form of instructions to improve perfor-
mance. Corrective feedback functions as an antecedent that evokes the correct
behavior in the next rehearsal so that it can be reinforced.

The best way to teach a skill is to provide instructions or modeling and to
require that the person rehearse the skill so that it can be reinforced. Although
instructions or modeling alone can evoke the correct behavior in the right situa-
tion, the behavior is not likely to continue to occur unless it is subsequently
reinforced. For example, suppose your friend told you to drive in the left lane
past the mall because cars that are about to turn into the mall slow the traffic in
the right lane. This is an instruction. You follow the instruction, and your behav-
ior is reinforced by avoiding slower traffic. As a result, you are more likely to drive
in the left lane past the mall. However, if you followed your friend’s instruction
and drove in the left lane but the traffic was not faster in that lane, your behavior
would not be reinforced. Therefore, even though the instruction evoked
(prompted) the correct behavior initially, the behavior would not continue to
occur because it was not reinforced after it occurred. When teaching a skill,
we could evoke the correct behavior simply by modeling it or by providing
instructions for the learner. However, to be sure that the behavior has been
learned, we also have the learner rehearse it in the simulated training situation so
that we can reinforce the behavior. It is much more likely that the learner will
execute the behavior in the real situation if the learner has already executed the
behavior successfully in training.

Behavioral Skills Training in Groups

Sometimes BST procedures are used with groups of people who all need to learn
similar skills. For example, parent training might be implemented with a group of
parents who are all having difficulty with their children; assertiveness training
might be conducted with a group of people who have assertive skills deficits.
Group BST is most effective with small groups in which all members have a
chance to participate (Himle & Miltenberger, 2004). In group BST, the modeling
and instructions are presented for the entire group. Each group member then
rehearses the skill in a role-play and receives feedback from the trainer and from
other members of the group (Poche et al., 1988). In group training, as with
individual BST, each person rehearses the skill until it is performed correctly in a
variety of simulated situations.

Group BST has a number of advantages. First, it can be more efficient than
individual BST because instructions and modeling are presented to the whole

Antecedent Behavior

Outcome: Client is more likely to engage in the correct skill in the role-play context.

Role-play context, modeling, and instructions Rehearsal of the skill Feedback (praise for correct performance)

Consequence

226 Chapter 12

group. Second, each group member learns by watching other group members
rehearse the skills and receive feedback on their performances. Third, group
members learn by evaluating the performance of other group members and
providing feedback. Fourth, with a variety of group members participating in
role-plays, generalization may be enhanced. Finally, the magnitude of reinforce-
ment for successful rehearsal is increased when praise comes from other group
members, as well as from the trainer.

A disadvantage of group BST is that each person does not have the trainer’s
undivided attention. One other possible problem is that some members may not
participate actively or may dominate and limit the participation of other members.
The trainer can prevent this problem by taking an active role and promoting
participation by all members.

Applications of Behavioral Skills Training
Procedures

Numerous studies have demonstrated that BST procedures are effective in teach-
ing a variety of skills (Rosenthal & Steffek, 1991). These procedures have been
used extensively with children. We have already discussed the studies by
Poche and her colleagues. Other researchers have also used BST procedures to
teach abduction prevention and sexual abuse prevention skills to children
(Carroll-Rowan & Miltenberger, 1994; Johnson et al., 2005, 2006; Miltenberger
& Thiesse-Duffy, 1988; Miltenberger, Thiesse-Duffy, Suda, Kozak, & Bruellman,
1990; Olsen-Woods et al., 1998; Wurtele, Marrs, & Miller-Perrin, 1987; Wurtele,
Saslawsky, Miller, Marrs, & Britcher, 1986). In each of these studies, the children
learned the correct responses to dangerous situations through modeling and
instructions, rehearsed the self-protection skills in role-plays of dangerous situa-
tions, and received feedback on their performance. These researchers found that
the use of instructions and modeling without rehearsal and feedback was less
effective for teaching children self-protection skills. The children learned much
more when they had an opportunity to rehearse the skills and receive feedback
on their performances after the instructions and modeling. Abduction prevention
and sexual abuse prevention skills have also been taught to adults with intellectual
disability using the same BST approach (Haseltine & Miltenberger, 1990;
Lumley, Miltenberger, Long, Rapp, & Roberts, 1998; Miltenberger, Roberts
et al., 1999). In some cases, in situ training was used after BST to help children
or individuals with intellectual disability learn the skills and use them in naturalis-
tic situations (Johnson et al., 2005, 2006).

In other research, BST procedures have been used to teach fire emergency
skills to children. Jones and Kazdin (1980) taught young children to make emer-
gency phone calls to the fire department. Jones, Kazdin, and Haney (1981) taught
children the skills they needed to respond to home fires. They identified nine
different home fire emergencies and the correct fire safety responses for each
situation. In training, they simulated a fire in a bedroom and used instructions,
modeling, rehearsal, and feedback to teach the child the correct responses.
The trainer told the child the correct behaviors and showed the child what to do.

Behavioral Skills Training Procedures 227

When the child executed the behavior correctly, the trainer provided praise and
other reinforcers. If a child performed any part of the behavior incorrectly, the
trainer gave feedback about what the child could do better, and the child
tried again until he or she did it right (Figure 12-3). Whenever any part of the
performance was incorrect, the researchers always praised the child for any portion
of the fire safety behavior that the child got right before providing the correction.
Their results are summarized in Figure 12-4.

BST procedures have also been used extensively with people who have social
skills deficits. For example, Elder, Edelstein, and Narick (1979) taught aggressive
adolescents to improve their social skills in an effort to reduce their aggressive
behavior. Matson and Stephens (1978) taught patients with chronic psychiatric
disorders to increase appropriate social behaviors, which resulted in a decrease in
arguing and fighting. Starke (1987) used BST procedures to improve the social
skills of physically disabled young adults. Warzak and Page (1990) taught sexually
active adolescent girls how to refuse unwanted sexual advances from adolescent
boys. In each study, the subjects learned the social skills through instructions and
modeling, rehearsal of the skills in role-plays, and feedback (reinforcement and
correction) on their performance.

Starke (1987) found that the BST procedure was more effective than a discus-
sion group for increasing social skills. This finding suggested that rehearsal and feed-
back were important components of the skills training procedure. In other words, it
is not enough to be told what skills are important and to see the skills demonstrated.
The best way to learn skills is to also have the opportunity for rehearsal and
feedback so that the skills can be reinforced in simulated or real situations.

Similar results showing that instructions and modeling are not sufficient, and
that learners must practice the skills with feedback to be successful, have been

FIGURE 12-3 The child is rehearsing a fire safety skill after viewing a model and receiving instructions from the
trainer. After the rehearsal, the trainer will provide feedback.

228 Chapter 12

Pe
rc

en
ta

ge
o

f C
or

re
ct

E
m

er
ge

nc
y

Re
sp

on
se

s

0

80

100

20

40

60

0 5 10 15 2 weeks20
Sessions

Don

• • •

Baseline Training Follow-Up

0

80

100

20

40

60

John

0

80

100

20

40

60

Dana

0

80

100

20

40

60

Lisa

0

80

100

20

40

60

Joe

FIGURE 12-4 This graph shows the percentage of correct fire emergency responses from five children before
and after behavioral skills training was implemented with each child. All of the children learned
the skills as a result of training. This graph is from a multiple-baseline-across-subjects design.
The performance of each child improved only after that child received training. (From Jones,
R. T., Kazdin, A. E., & Haney, J. L. [1981]. Social validation and training of emergency fire safety
skills for potential injury prevention and life saving. Journal of Applied Behavior Analysis, 14,
249–260. Copyright © 1981 University of Kansas Press. Reprinted by permission of the publisher.)

Behavioral Skills Training Procedures 229

reported by researchers (Beck & Miltenberger, 2009; Gatheridge et al., 2004;
Himle, Miltenberger, Gatheridge et al., 2004; Poche et al., 1988). For example, in
the study by Beck and Miltenberger, children watched a video purchased from the
internet designed to teach abduction prevention skills to children (say “no,” run
away, and tell a parent when approached by a stranger). Although the video was
highly regarded and won a number of awards for its quality, after the children
watched the video, they were not able to engage in the abduction prevention skills
during an in situ assessment (when they were approached by a stranger in a store
without their knowledge that they were being tested). However, after receiving in
situ training where they practiced the abduction prevention skills and received feed-
back, they were successful in using the skills in further assessments. This has been a
consistent finding in the research—telling and showing kids what to do is not
enough; they have to practice the skills with feedback (reinforcement and error cor-
rection) in order to use the skills in actual situations where the skills are needed.

Finally, researchers have demonstrated that BST procedures are effective in
teaching skills to adults. Forehand and his colleagues (Forehand et al., 1979) used
these procedures to teach child management skills to parents of noncompliant chil-
dren. The parents learned skills needed to reward their children, make requests appro-
priately, and use time-out when their children were noncompliant. When the parents
learned these skills, their children’s behavior improved. Other researchers have shown
that BST procedures are effective in teaching behavior modification skills to teachers
or staff who work with children, nursing home residents, or individuals with intellec-
tual disability (e.g., Engelman, Altus, Mosier & Mathews, 2003; Lavie & Sturmey,
2002; Moore et al., 2002; Sarokof & Sturmey, 2004). Miltenberger and Fuqua
(1985b) used instructions, modeling, rehearsal, and feedback to teach college students
how to conduct clinical interviews. The students learned to ask the right kinds of
questions when conducting an interview with research assistants who were simulating
clients with behavior problems. Dancer and his colleagues (Dancer et al., 1978)
taught behavioral observation and description skills to married couples who were
going to manage group homes for delinquent youths. The couples needed these skills
to work effectively with the youths, who exhibited a variety of behavior problems.

The research cited here is just a sample of the applications of BST proce-
dures. Such procedures are used with people who can learn from instructions
and modeling in simulated situations and do not need the intensive training pro-
vided in the chaining procedures described in Chapter 11. Chaining procedures
generally are used with people who have limited abilities and need intensive
prompting. BST procedures, by contrast, often are used with children and adults
with normal abilities. However, they have also been used with people with disabil-
ities. For example, Hall, Sheldon-Wildgen, and Sherman (1980) used instructions,
modeling, rehearsal, and feedback to teach job interview skills to adults with mild
or moderate disabilities. After describing and modeling the important verbal and
nonverbal skills in an interview, Hall and her colleagues had the learners rehearse
the skills in simulated interviews.

What do you think Hall did after the rehearsals in the simulated interviews?

After the rehearsal, Hall provided praise for appropriate behaviors and
described the behaviors that the learners needed to improve.

230 Chapter 12

Using BST procedures, Bakken, Miltenberger, and Schauss (1993) taught
parents with intellectual disabilities important skills for interacting with their
children. The parents learned to praise their children and pay attention to them
in appropriate ways to facilitate normal development. An interesting finding from
this study was that the parents learned the skills when instructions, modeling,
rehearsal, and feedback were used in training sessions, but the skills did not gener-
alize to everyday situations in the home. When Bakken implemented training in
the home, the parents started to exhibit the skills there also. This finding under-
scores the importance of assessing the generalization of skills to the natural settings
where the skills are needed and providing further training if generalization does
not occur. (See Chapter 19 for further discussion of generalization.)

FOR FURTHER READING
Using Behavioral Skills Training to Teach Safety Skills to Prevent Gunplay

When a child finds a gun that has been left unattended by an adult, the child often plays with the gun
(e.g., Himle et al., 2004). As a result, the child might accidentally injure or kill himself or herself or
another person if the gun discharges. In response to this problem, researchers have evaluated BST for
teaching safety skills to children to prevent gunplay (e.g., Himle, Miltenberger, Flessner, & Gatheridge,
2004; Miltenberger et al., 2004, 2005). The safety skills a child is taught to use when finding a gun
are: (a) don’t touch the gun, (b) get away immediately, and (c) tell an adult. Researchers have shown
that BST can be successful in teaching 4- to 7-year-old children these skills, but that, in some cases,
in situ training is needed. In these studies, the researchers used in situ assessments in baseline and
after training—they set up situations in which a child found a gun (a real but disabled handgun) without
knowing that he or she was being assessed. To conduct in situ training, a researcher observed the post-
training assessment without being seen by the child, and if the child did not use the safety skills when
finding the gun, the researcher walked into the room and asked the child what he or she should have
done when finding the gun. After the child described the correct behaviors, the researcher then required
the child to practice the skills five times in the situation where the child found the gun. The researchers
found that all children learned the skills with BST and in situ training.

How to Use Behavioral Skills Training
Procedures

The following steps outline the effective use of BST procedures.

1. Identify and define the skills you want to teach. A good behavioral defini-
tion will clearly describe all the behaviors involved in the skills. You
should define all the skills that may be needed in various situations and
conduct a task analysis of complex skills (behavioral chains).

2. Identify all relevant stimulus situations (SDs) in which the skills must be
used. For example, in teaching abduction prevention skills, you have to
identify all possible abduction lures that a person might use so that the
child can learn to respond successfully to every abduction situation.
When teaching assertiveness, you have to identify all possible situations
in which a person might act unassertively so that the person can learn to
respond assertively in every situation.

Behavioral Skills Training Procedures 231

3. Assess the learner’s skills in the stimulus situations to establish a baseline.
To assess the learner’s skills, you must present each stimulus situation
(either the real situation or a simulation) and record the learner’s
response to that situation.

4. Begin training with the easiest skill or the easiest stimulus situation.
In those circumstances, the learner is more likely to be successful in
training and is more likely to continue cooperating with the BST proce-
dure. If you start with more difficult skills or situations, the learner may
not be successful initially and may become discouraged.

5. Begin a training session by modeling the behavior and describing its
important aspects. Be sure to model the behavior in the proper context
(in response to the relevant SD). You can create the proper context by
simulating it in a role-play. The simulation should be as real as possible
for the learner. Sometimes training sessions are conducted in the real
environment; for example, Poche and her colleagues (1981) modeled
abduction prevention skills out on the playground, where a child might
actually be approached by a potential abductor.

6. After the learner hears the instructions and sees the model, provide the
opportunity for rehearsal. Simulate the proper context for the behavior
and have the learner practice the behavior. Sometimes the simulation
or role-play may occur in the natural situation. Poche and her collea-
gues (1981) had children rehearse abduction prevention skills on the
playground.

7. Immediately after the rehearsal, provide feedback. Always provide descrip-
tive praise for some correct aspect of the performance. Then provide
instructions for improvement as needed.

8. Repeat the rehearsal and feedback process until the learner has executed
the behavior correctly a couple of times.

9. After success with one training situation, move to another situation and
continue the process of modeling, instructions, rehearsal, and feedback
until the learner has mastered each skill in each situation. While adding
new situations, continue to have learners practice training situations they
have mastered to ensure maintenance.

10. Once the learner has mastered all the skills in all simulated situations
during training sessions, program for generalization to the natural situa-
tions where the skills are needed. If the training situations are as similar
as possible to the natural situations, or if training occurs in the natural
situation (e.g., Poche et al., 1981), generalization is more likely to
occur. Another way to enhance generalization is to have the learner
practice the skills in progressively more difficult situations. For example,
after training social skills, you give the learner instructions to use the
social skills in real situations with real people in the learner’s life.
Start with easy assignments and, as the learner is successful, work up to
harder ones. The key point is to maintain success so that the
learner’s efforts are reinforced. Other ways to promote generalization
are reviewed in Chapter 19.

232 Chapter 12

CHAPTER SUMMARY

1. Behavioral skills training (BST) procedures consist of
four components: modeling, instructions, rehearsal,
and feedback. These training components have
been used together to teach a variety of important
skills to people with disabilities and to a wide range
of other adults and children. First, the trainer pro-
vides live or symbolic modeling so that the learner
sees how to execute the behavior. The trainer also
provides instructions in which the important aspects
of the behavior are described for the learner. The
learner then gets the opportunity to rehearse the
behavior in a simulated situation similar to the
natural situation in which the behavior is needed.
After the rehearsal, the trainer provides feedback
consisting of reinforcement for correct aspects of
the behavior and instructions on how to improve
the behavior. Further rehearsals are conducted,
and feedback is provided, until the learner displays
the correct behavior in a variety of relevant contexts.

2. The appropriate time to use BST procedures is
when the learner can benefit from modeling and
instructions and does not need more intensive train-
ing procedures (such as chaining procedures) to
learn the skills.

3. You conduct BST in small groups by providing
modeling and instruction for the whole group
and then having each member of the group indi-
vidually rehearse the skills in role-plays and
receive feedback. Feedback may come from the
trainer, as well as from other group members.

4. BST procedures involve a three-term contingency
for the skill being learned. Modeling and instruc-
tions are antecedents to get the correct behavior to
occur, the correct behavior occurs in a rehearsal,
and feedback is provided as a reinforcing conse-
quence for the behavior in the rehearsal. Feedback
may also involve further instructions that act as a
prompt for the behavior in the next rehearsal.

KEY TERMS

behavioral skills training
procedures (BST), 217

feedback, 223

in situ assessment, 224
in situ training, 225
instructions, 221

modeling (modeling
prompt), 218

rehearsal, 222

PRACTICE TEST

1. What four procedures are components of the
BST procedure? Describe each component
procedure. (pp. 218–223)

2. Describe the use of the BST procedure.
(pp. 217–218)

3. Provide two examples (not from this chapter)
of skills that could be taught through the BST
procedure.

4. For both of these examples, describe how you
would use the BST procedure. (pp. 218–223)

5. Why is the use of instructions or modeling alone
usually not effective in the long run? (p. 226)

6. Describe the factors that enhance the effective-
ness of modeling. What factors reduce the effec-
tiveness of modeling? (pp. 218–221)

7. Describe the factors that influence the effective-
ness of instructions. (pp. 221–222)

8. When using rehearsal, why should you start with
easy behaviors or situations? What might happen
if you practiced the most difficult situations first?
(p. 222)

9. Describe the factors that influence the effective-
ness of rehearsal. (pp. 222–223)

10. Describe the two types of feedback you
can provide after a behavioral rehearsal.
(p. 223)

11. When providing feedback after a behavioral
rehearsal, why should you always provide praise
first? What should you do if the behavior was not
correct in the rehearsal? (p. 223)

12. Describe the factors that influence the effective-
ness of feedback. (p. 223)

13. Describe how the three-term contingency is
involved in the BST procedure. (pp. 225–226)

Behavioral Skills Training Procedures 233

14. Describe the guidelines for the effective use of
the BST procedure. (pp. 231, 232)

15. How is the BST procedure different from the
chaining procedures described in Chapter 11?
How are they similar? (pp. 231–232)

16. In what circumstances would a chaining proce-
dure be most appropriate? In what circum-
stances would the BST procedure be most
appropriate? (p. 230)

APPLICATIONS

1. You are a school counselor and you have been
asked to teach a group of eighth graders the skills
they will need to resist peer pressure to start
smoking. Describe how you will use BST proce-
dures to teach these kids these important skills.
Assume that you will work with groups of 20–25
kids in each classroom.
a. Define the skills you will teach.
b. Identify the situations in which the kids will

need these skills.
c. Create the role-plays you will use in training.
d. Describe how you will model the behavior

and what instructions you will give.
e. Describe the types of rehearsal and feedback

you will use.
f. Describe what you will do to increase the

chances for generalization of the skills the
kids will learn.

2. Your young daughter is in first grade and she
wants to walk to school (two blocks away) with
her friends every day. You have decided that she
must learn some personal safety skills before you

will allow her to walk to school without adult
supervision. You want to teach her how to
respond to an adult who offers her a ride to or
from school. You do not want her to accept a
ride from anybody without your permission.
Describe the BST procedure you will use to
teach her the skills she will need to respond
safely in such a situation. Address each of the
points raised in Application 1. Also, describe
how you will assess her skills after training to
be sure that the skills have generalized to the
natural situation.

3. You are teaching a class of ten parents who are
having trouble with their children. All the par-
ents have a child who engages in attention-
seeking behavior such as whining, crying, or
interrupting. One of the things you want to
teach the parents is how to differentially rein-
force their child’s appropriate behavior, such as
playing or doing a chore. Describe how you will
use the BST procedure to teach the parents how
to reinforce their child’s good behavior.

MISAPPLICATIONS

1. The principal of an elementary school has
decided it is time to teach the students about
drugs and how to resist someone who offers
them drugs or tries to talk them into trying or
selling drugs. The principal gets a film that talks
about the dangers of drugs and tells kids never to
take drugs or sell drugs. The film repeats the
message that kids should just say no and walk
away from a person with drugs. The film shows
a few kids saying no and walking away. The
principal shows the film in each classroom and
asks the kids whether they have any questions.

What is the problem with the principal’s plan for
teaching the students to say no to drugs? How
would you improve on this plan?

2. After supper each day, workers at a group home for
adolescents with intellectual disabilities are sup-
posed to conduct training programs in toothbrush-
ing,grooming, housekeeping,and otherskills. The
supervisor goes home at 5 P.M., and the staff mem-
bers often sit around and talk after supper instead of
training the residents. Whenever the supervisor
drops by, staff members get up and do their jobs,
but they stop working again when he leaves.

234 Chapter 12

The supervisor decides to conduct BST with the
staff. He conducts a few training sessions in which
he uses modeling, instruction, rehearsal, and feed-
back to teach staff members the skills they need to
work with the residents. He believes that as a result
of the training, the staff will use these skills when
he is not present. What is wrong with this use of
the BST procedure? What would be a more appro-
priate procedure?

3. In a new campaign, major sports stars in TV
commercials tell kids to stay in school, study
hard, and get good grades. The commercials tar-
get inner-city high school and junior high kids.
The sports stars tell kids why they should study

and how it will improve their lives in the future.
The commercial shows some kids studying and
older kids telling them how smart they are for
studying. It shows kids refusing to go out at night
with other kids because they have to study. After
the kids model this behavior, the sports star
praises them and says how smart it is to study
and stay in school. Finally, the commercial
shows kids graduating and getting good jobs.
Again, the sports star comes on-screen and
points out the good things that studying and stay-
ing in school can get them. What is good about
this strategy to encourage kids to study? What is
missing? How could you improve this strategy?

Behavioral Skills Training Procedures 235

This page intentionally left blank

Chapter Thirteen

Understanding Problem Behaviors
through Functional Assessment

Chapters 9 through 12 describe procedures for establishing desirable behaviors.This section of the text describes behavioral procedures for understanding
problem behaviors and increasing or decreasing existing behaviors. When using
behavior modification procedures to help a person increase a desirable behavior
or decrease or eliminate an undesirable behavior (a problem behavior), the first
step is to understand why the person engages in the behavior. To do so, you
must conduct an assessment of the three-term contingency to determine the ante-
cedent events that evoke the behavior and the reinforcing consequences that
maintain it. The process of identifying these variables before treating a problem
behavior is called functional assessment.

Examples of Functional
Assessment

Jacob
Jacob, a 2-year-old boy, lived with his mother and his 4-year-old sister.
His mother ran a day care business out of her house and took care of
10-15 other young children. Jacob engaged in problem behaviors involv-
ing throwing objects, banging his head on the ground, and whining. His
mother was concerned about Jacob’s problems and agreed to participate
in a behavior modification experiment, conducted by a psychology
graduate student named Rich, to try to decrease Jacob’s problem beha-
viors (Arndorfer, Miltenberger, Woster, Rortvedt, & Gaffaney, 1994).
The first step Rich took was to conduct a functional assessment to
determine why Jacob was engaging in these behaviors.

First, Rich interviewed Jacob’s mother and asked her questions
about the problem behaviors, the setting and the day care routines, the antecedent
circumstances, the consequences when Jacob engaged in the problem behaviors,
other behaviors that Jacob engaged in, and previous treatments that she had tried
with Jacob. After the interview, Rich observed Jacob in the day care setting and
recorded information on the antecedents, behavior, and consequences each time

â–  What is a functional assessment of
a problem behavior?

â–  What are the three ways to conduct
a functional assessment?

â–  How do you use indirect methods to
conduct a functional assessment?

â–  How do you use direct observation
methods to conduct a functional
assessment?

â–  What is a functional analysis of
a problem behavior? How do you
conduct a functional analysis?

237

Jacob engaged in the problem behaviors. He observed Jacob for a few days until
he could determine which antecedents and consequences were reliably associated
with the behavior.

On the basis of the information from the interview and the observations, Rich
developed a hypothesis about the function of the problem behaviors. He deter-
mined that Jacob was more likely to engage in the problem behaviors when other
children in day care took his toys or tried to play with his toys. Furthermore, when
Jacob engaged in the head-banging, whining, or toy-throwing, the other children
were likely to stop playing with his toys and give the toys back to him. Rich
hypothesized that the reinforcer for the problem behaviors was that the other
children gave Jacob back his toys.

To determine whether this hypothesis was correct, Rich conducted a brief
experiment. On some days, he instructed the other children in day care not to
touch Jacob’s toys; on other days, he instructed the children to play with Jacob’s
toys but to give the toys back to him immediately if he engaged in the problem
behaviors. Rich found that Jacob was much more likely to engage in the prob-
lem behaviors on days when the other children played with his toys. On days
when the other children did not touch his toys, Jacob rarely engaged in the
problem behaviors. The brief experiment had confirmed that other children
playing with Jacob’s toys was an antecedent for the problem behaviors. Further-
more, it confirmed that the reinforcer for the problem behaviors was that the
other children gave back the toys.

Treatment for Jacob involved teaching him to ask the other children to give
back his toys when they took them. Asking for the toys is a behavior that is func-
tionally equivalent to the problem behaviors. In other words, asking for the toys
produced the same outcome as the problem behavior: The children gave the
toys back to him. When Jacob exhibited aggressive behavior, he did not get his
toys back.

The treatment helped Jacob replace the undesirable behavior (head-banging,
whining, toy-throwing) with a desirable behavior (asking for the toys). This

Antecedent Behavior

Outcome: Jacob is more likely to engage in head-banging, whining, and toy-throwing when other children play with his toys.

Other kids play with Jacob’s toys. Jacob bangs his head, whines, and throws toys. The kids return Jacob’s toys to him.

Consequence

Antecedent Behavior

Outcome: Jacob is more likely to ask for his toys back when other children play with them.

Other kids play with Jacob’s toys. Jacob asks for his toys back. The kids return Jacob’s toys to him.

Consequence

238 Chapter 13

treatment strategy, using differential reinforcement to increase a desirable behavior
and decrease an undesirable behavior, is described in Chapter 15. The functional
assessment Rich conducted with Jacob helped him choose an effective treatment
for Jacob’s problem behaviors. Conducting a functional assessment is always the
first step in using behavior modification procedures to decrease problem behaviors.

Anna
Anna, a 3-year-old girl, lived with her mother and younger sister. Anna engaged in
problem behaviors at home involving hitting, kicking, and screaming (Arndorfer et
al., 1994). To understand the function of these behaviors, Rich again conducted a
functional assessment. He interviewed Anna’s mother and then conducted direct
observations of the three-term contingency related to the problem behaviors. On
the basis of the results of the interview and his observations, Rich hypothesized
that Anna’s problem behaviors were reinforced by her mother’s attention. Anna
was most likely to engage in problem behaviors when her mother was not paying
attention to her (e.g., when her mother was working around the house). Further-
more, the most common consequence of the problem behavior was that Anna’s
mother immediately stopped what she was doing and paid attention to Anna.
Rich conducted a brief experiment to confirm this hypothesis.

What do you think Rich did in his brief experiment?

Rich had Anna’s mother manipulate her level of attention to Anna to determine
whether her attention reinforced Anna’s problem behaviors. In the first condition, she
played with Anna and paid attention to her. If Anna engaged in a problem behavior,
her mother ignored it. In the second experimental condition, she paid little attention
to Anna and focused on a task instead. If Anna exhibited a problem behavior, her
mother immediately stopped what she was doing and paid attention to Anna for a
brief period of time. Rich found that Anna exhibited a much higher frequency of the
problem behaviors in the second condition. This confirmed the hypothesis that
the reinforcer for Anna’s problem behavior was her mother’s attention.

Rich implemented a treatment similar to the one he used with Jacob.
He taught Anna how to ask for her mother’s attention when her mother was not
paying attention to her. He taught Anna’s mother to differentially reinforce
Anna’s behavior of asking for attention. That is, when Anna asked for attention,
her mother immediately paid attention to her for a brief period of time. However,
when Anna engaged in the problem behavior, her mother used extinction and did
not provide any attention.

Antecedent Behavior

Outcome: Anna is more likely to hit, kick, and scream when her mother is not paying attention to her.

Anna’s mother is not paying attention to her. Anna hits, kicks, and screams. Anna’s mother pays attention to her.

Consequence

Understanding Problem Behaviors through Functional Assessment 239

When Anna engaged in a problem behavior, her mother’s only reaction was
to take Anna’s little sister into the other room with her so that the little sister
would not get hurt (because Anna’s problem behavior involved hitting and kick-
ing). Rich found that the use of differential reinforcement resulted in a decrease
in the problem behavior and an increase in the desirable behavior of asking her
mother for attention. Once again, the particular treatment chosen for Anna was
based on information from the functional assessment conducted as the first step
in the treatment process.

Note that sometimes, when a child learns to ask for attention as an alternative
to the problem behavior, the child may then ask for attention so often that this
behavior itself becomes a problem. Carr and his colleagues (Carr et al., 1994)
have described procedures to solve this problem. With each successive request for
attention, the parents wait longer and longer before responding. Eventually, the
child asks less often.

Defining Functional Assessment

One basic principle of behavior analysis is that behavior is lawful. Regardless of
whether the behavior is desirable or undesirable, its occurrence is controlled by
environmental variables; that is, the behavior occurs as a function of environmen-
tal variables. Respondent behavior is controlled by antecedent stimuli, and operant
behavior is controlled by antecedents and consequences that make up three-term
contingencies of reinforcement and punishment. Functional assessment is the pro-
cess of gathering information about the antecedents and consequences that are
functionally related to the occurrence of a problem behavior. It provides informa-
tion that helps you determine why a problem behavior is occurring (Drasgow,
Yell, Bradley, & Shiner, 1999; Ellis & Magee, 1999; Horner & Carr, 1997;
Iwata, Vollmer, & Zarcone, 1990; Iwata, Vollmer, Zarcone, & Rodgers, 1993;
Larson & Maag, 1999; Lennox & Miltenberger, 1989; Neef, 1994).

In addition to information on the reinforcing consequences (functions) of
target behaviors, a functional assessment also provides detailed information about
antecedent stimuli, including the time and place of the behavior, people present
when the behavior occurs, any environmental events immediately preceding
the behavior, and the frequency (or other dimensions) of the target behavior.
This information on the three-term contingency will help you to identify the
antecedents that have stimulus control over the behavior and the reinforcing
consequences that maintain the behavior.

Functional assessment also provides other types of information that are impor-
tant for developing appropriate treatments for problem behaviors, including the

Antecedent Behavior

Outcome: Anna is more likely to ask for her mother’s attention when her mother is not paying attention to her at the time.

Anna’s mother is not paying attention to her. Anna asks her mother for attention. Anna’s mother pays attention to her.

Consequence

240 Chapter 13

existence of alternative behaviors that may be functionally equivalent to the
problem behavior, motivational variables (establishing operations and abolishing
operations that influence the effectiveness of stimuli as reinforcers and punishers),
stimuli that may function as reinforcers for the person, and the history of previous
treatments and their outcomes (Table 13-1).

Functions of Problem Behaviors

A primary purpose of a functional assessment is to identify the function of the
problem behavior. There are four broad classes of reinforcing consequences or
functions of problem behaviors (Iwata et al., 1993; Miltenberger, 1998, 1999).

Social Positive Reinforcement
One type of reinforcing consequence involves positive reinforcement mediated by
another person. When a positively reinforcing consequence is delivered by
another person after the target behavior, it is called social positive reinforcement.
Social positive reinforcement may involve attention, access to activities, or tangi-
bles provided by another person. For example, Anna received attention from her
mother as a reinforcer for her problem behavior, and Jacob received his toys back
from the other kids (tangibles) as a reinforcer for his problem behavior. In both
cases, these consequences make the behavior more likely to occur.

Social Negative Reinforcement
In some cases, target behaviors are maintained by negative reinforcement that is
mediated by another person. When another person terminates an aversive interac-
tion, task, or activity after the occurrence of a target behavior, and as a result, the

TABLE 13-1 Categories of Information from a Functional Assessment

â–  Problem behaviors: an objective description of the behaviors that make up the problem
â–  Antecedents: an objective description of environmental events preceding the problem behavior, includ-

ing aspects of the physical environment and the behavior of other people

â–  Consequences: an objective description of environmental events that follow the problem behavior,
including aspects of the physical environment and the behavior of other people

■ Alternative behaviors: information on desirable behaviors in the person’s repertoire that may be rein-
forced to compete with the problem behavior

â–  Motivational variables: information on environmental events that may function as establishing opera-
tions or abolishing operations to influence the effectiveness of reinforcers and punishers for the prob-
lem behaviors and alternative behaviors

■ Potential reinforcers: information on environmental events—including physical stimuli and the behav-
ior of other people—that may function as reinforcers and be used in a treatment program

â–  Previous interventions: information on the interventions that have been used in the past and their
effects on the problem behavior

Understanding Problem Behaviors through Functional Assessment 241

behavior is more likely to occur, the behavior is said to be maintained by social neg-
ative reinforcement. For example, a child who complains to his parent when asked
to do a chore may get out of doing the chore as a result of complaining. Likewise, a
student who bangs her head when instructed to do an academic task may escape
from the task as a result. In each case, being allowed to escape from the chore or
task strengthens or reinforces the problem behavior. Asking a friend not to smoke
in your car is negatively reinforced by escape or avoidance of the smell of the
smoke when the person puts out the cigarette or does not light it in the first place.

Automatic Positive Reinforcement
In some cases, the reinforcing consequence of a target behavior is not mediated by
another person, but rather occurs as an automatic consequence of the behavior
itself. When the behavior produces a reinforcing consequence automatically, and
the behavior is strengthened, the behavior is said to be maintained by automatic
positive reinforcement. For example, some behaviors produce sensory stimulation
that reinforces the behavior. A child with autism who spins objects, rocks in his
seat, or flaps his fingers in front of his face may do so because the behaviors pro-
duce reinforcing sensory stimulation. In this case, the reinforcing consequence for
the behavior is not mediated by another person. Going to the kitchen to get a
drink is automatically positively reinforced by getting the drink, whereas asking
someone else to get you a drink is socially positively reinforced by getting the
drink from the other person.

Automatic Negative Reinforcement
Automatic negative reinforcement occurs when the target behavior automatically
reduces or eliminates an aversive stimulus as a consequence of the behavior and
the behavior is strengthened. With automatic negative reinforcement, escape
from the aversive stimulus is not mediated by the actions of another person.
Closing the window to block a cold draft involves automatic negative reinforce-
ment. Asking someone to close the window to get rid of the draft involves social
negative reinforcement. An example of a problem behavior that may be main-
tained by automatic negative reinforcement is binge eating. In some cases, binge
eating has been found to be maintained by the reduction in unpleasant emotional
responses that were present before binge eating (Miltenberger, 2005; Stickney &
Miltenberger, 1999; Stickney, Miltenberger, & Wolff, 1999). That is, when the
person experiences strong unpleasant emotions, binge eating temporarily decreases
the unpleasant emotions, thus negatively reinforcing binge eating.

Functional Assessment Methods

The various methods used to conduct functional assessments fall into three categories:
indirect assessment methods, in which information is gathered through interviews and
questionnaires; direct observation methods, in which an observer records the antece-
dents, behavior, and consequences as they occur; and experimental methods (also
called functional analysis), in which antecedents and consequences are manipulated

242 Chapter 13

to observe their effect on the problem behavior (Iwata, Vollmer, & Zarcone, 1990;
Lennox & Miltenberger, 1989). Let’s consider each of these approaches in turn.

Functional Assessment Methods
â–  Indirect methods

â–  Direct observation methods

â–  Experimental methods (functional analysis)

Indirect Methods
With indirect functional assessment methods, behavioral interviews or questionnaires
are used to gather information from the person exhibiting the problem behavior
(the client) or from others who know this person well (e.g., family members, tea-
chers, or staff). Indirect assessment methods are also known as informant assessment
methods because an informant (the client or others) is providing information in
response to assessment questions (Lennox & Miltenberger, 1989). The advantage of
indirect functional assessment methods is that they are easy to conduct and do not
take much time. In addition, a number of interview formats and questionnaires are
available for use in conducting a functional assessment (Bailey & Pyles, 1989;
Durand & Crimmins, 1988; Iwata, Wong, Riordan, Dorsey, & Lau, 1982; Lewis,
Scott, & Sugai, 1994; Miltenberger & Fuqua, 1985b; O’Neill, Horner, Albin,
Storey, & Sprague, 1990; O’Neill et al., 1997). The disadvantage of indirect methods
is that the informants must rely on their memory of the events. Thus, information
from interviews and questionnaires may be incorrect as a result of forgetting or bias.

Because of their convenience, indirect functional assessment methods are used
commonly. In fact, the interview is the most common assessment method used
by psychologists (Elliott, Miltenberger, Bundgaard, & Lumley, 1996; Swan &
MacDonald, 1978). A good behavioral interview is structured to generate information
from the informant that is clear and objective. Information about the problem behavior,
antecedents, and consequences should describe environmental events (including the
behavior of other people) without inferences or interpretation. For example, consider
two different answers to the interview question, “When does your child engage in the
tantrum behavior?” (Assume that tantrum behavior has already been described by the
parent.) If the parent says, “Johnny has a tantrum when I tell him to turn off the TV
and come to the dinner table,” the parent is providing objective information about envi-
ronmental events that immediately precede the problem. If the parent says, “Johnny has
a tantrum when he doesn’t get to do what he wants,” the parent is interpreting the situa-
tion. This second answer does not provide objective information about the antecedents
of the problem. It does not describe specific environmental events.

The goal of a behavioral interview is to generate information on the problem
behaviors, antecedents, consequences, and other variables that will permit you to
form a hypothesis about the controlling variables for the problem. At the same
time, an effective interview teaches the client or informant about functional assess-
ment: that behaviors and events must be identified and specified, that inferences
should be minimized, and that it is important to focus on antecedents and
consequences in understanding and changing behavior. The following is a list of

Understanding Problem Behaviors through Functional Assessment 243

questions an interviewer might ask to generate information about the antecedents
and consequences of a child’s problem behavior.

Antecedents
â–  When does the problem behavior usually occur?

â–  Where does the problem behavior usually occur?

â–  Who is present when the problem behavior occurs?

â–  What activities or events precede the occurrence of the problem behavior?

â–  What do other people say or do immediately before the problem behavior occurs?

â–  Does the child engage in any other behaviors before the problem behavior?

â–  When, where, with whom, and in what circumstances is the problem behavior least likely to occur?

Consequences
â–  What happens after the problem behavior occurs?

â–  What do you do when the problem behavior occurs?

â–  What do other people do when the problem behavior occurs?

â–  What changes after the problem behavior occurs?

â–  What does the child get after the problem behavior?

â–  What does the child get out of or avoid after the problem behavior?

Each of these questions asks about the events that immediately precede and fol-
low the child’s problem behavior. The interviewer asks these questions in the hope
that the parent will provide objective information. If the parent does not provide
specific information about environmental events in response to one or more ques-
tions, the interviewer will ask for clarification until the parent provides information
that shows a clear pattern of events that precede and follow the problem behavior.
Once the interviewer can discern a reliable pattern of antecedents and conse-
quences, the interviewer can develop a hypothesis about the antecedents that have
stimulus control over the problem behavior and the reinforcer that maintains it.

Various authors have developed lists of questions to generate thorough functional
assessment information in a behavioral interview. Table 13-2 shows the categories of
assessment information and sample interview questions from the Functional Analysis
Interview Format developed for use with staff, teachers, and others who work with
people with intellectual disabilities (O’Neill et al., 1990, 1997). These questions can
be answered in an interview or questionnaire format (Ellingson, Miltenberger,
Stricker, Galensky, & Garlinghouse, 2000; Galensky, Miltenberger, Stricker, & Gar-
linghouse, 2001). In an interview format, the interviewer asks the informant each
question and records the answer. In a questionnaire format, the informant reads each
question and writes down the answer. If the questions are used in a questionnaire for-
mat, the professional reviews the answers and then follows up with an interview to
clarify any answers that did not provide complete or objective information.

Because indirect functional assessment methods have the disadvantage of relying
on informants’ memories of events, researchers suggest using multiple functional

244 Chapter 13

assessment methods to produce the most accurate information on antecedents, conse-
quences, and the other variables listed in Table 13-1 (Arndorfer & Miltenberger, 1993;
Arndorfer et al., 1994; Ellingson et al., 2000). Arndorfer and his colleagues suggest
that a behavioral interview combined with direct observation of the antecedents and
consequences provides useful information that enables you to formulate accurate
hypotheses about the function of the problem behavior.

Direct Observation Methods
When conducting a functional assessment using direct observation methods, a
person observes and records the antecedents and consequences each time the
problem behavior occurs. The person conducting the direct observation

TABLE 13-2 Categories of Assessment Information and Sample Questions from the Functional
Analysis Interview Format

A. Describe the behaviors.

â–  What are the behaviors of concern?

â–  For each behavior, define how it is performed, how often it occurs, and how long it lasts.

B. Define potential ecological events that may affect the behaviors.

â–  What medications is the person taking, and how do you think these may affect the behaviors?

â–  How many other people are in the setting (work/school/home)? Do you believe that the density of peo-
ple or interactions with other people affect the targeted behaviors?

â–  What is the staffing pattern? To what extent do you believe the number of staff, training of staff, and
quality of social contact with staff affect the targeted behaviors?

C. Define events and situations that predict occurrences of the behaviors (antecedents).

â–  When, where, and with whom are behaviors most likely? Least likely?

â–  What activity is most likely to produce the behaviors? Least likely?

D. Identify the function of the undesirable behaviors. What consequences maintain the behaviors?

â–  What does the person get and what does the person avoid as a consequence of the behaviors?

E. Define the efficiency of the undesirable behaviors.

â–  What amount of physical effort is involved in the behaviors?

â–  Does engaging in the behaviors result in a payoff every time?

F. Define the primary methods the person uses to communicate.

â–  What general expressive communication strategies does the person use?

G. Identify potential reinforcers.

â–  In general, what factors (events/activities/objects/people) appear to be reinforcing or enjoyable for
the person?

H. What functional alternative behaviors does the person know?

â–  What socially appropriate behaviors or skills does the person perform that may be ways of
achieving the same function(s) as the behaviors of concern?

I. Provide a history of undesirable behaviors and the programs that have been attempted.

â–  Identify the treatment programs and how effective they have been.

Source: Adapted from O’Neill, Horner, Albin, Storey, & Sprague, 1990.

Understanding Problem Behaviors through Functional Assessment 245

assessment (the observer) may be the person exhibiting the problem behavior, or it
may be another person associated with the client, such as a parent, teacher, staff
person, nurse, behavior analyst, or psychologist. The antecedents and conse-
quences are observed and recorded in the natural environment where the prob-
lem behavior typically takes place. An exception would be when observations
occur while a person is in a treatment setting (e.g., a hospital or clinic). Direct
observation assessment also is called ABC observation. The goal of ABC observa-
tions is to record the immediate antecedents and consequences typically associ-
ated with the problem behavior under normal conditions (Anderson & Long,
2002; Bijou, Peterson, & Ault, 1968; Lalli, Browder, Mace, & Brown, 1993;
Repp & Karsh, 1994; Vollmer, Borrero, Wright, Van Camp, & Lalli, 2001).

Advantages and disadvantages are associated with ABC observation as a method
of conducting a functional assessment of a problem behavior. The main advantage of
ABC observations over indirect methods is that an observer is recording the antece-
dents and consequences as they occur rather than reporting the antecedents and
consequences from memory. The assessment information is likely to be more accu-
rate when it comes from direct observation. A disadvantage is that ABC observations
take more time and effort than interview or questionnaire methods. In addition,
even though ABC observations produce objective information about the antece-
dents and consequences that are reliably associated with the problem behavior,
ABC observations do not demonstrate a functional relationship, but rather a correla-
tion of the antecedents and consequences with the problem behavior. To demon-
strate that a functional relationship exists, experimental methods (functional
analysis) must be used; these methods are described in the next section. However,
even though the ABC observations demonstrate only a correlation of the antece-
dents and consequences with the problem behavior, the information allows you to
develop a hypothesis about the antecedents that evoke the behavior and the rein-
forcer that maintains the behavior. The development of a hypothesis about the
antecedents and consequences is the desired outcome of conducting ABC observa-
tions. A strong hypothesis about the controlling antecedents and consequences
often is sufficient to develop effective treatment strategies. Your hypothesis about
the controlling variables is strengthened when the information from indirect assess-
ments is consistent with information from the ABC direct observation assessment.

To conduct ABC observations, the observer should be present in the client’s
natural environment when the problem behavior is most likely to occur. For example,
if a student has problem behaviors in one class but not in others, the observer should
be present in that particular class to observe and record the ABCs. Therefore, to make
ABC observations most efficient, it is helpful to know in advance when the problem
behavior is most likely to occur. Information from an interview may indicate when
the problem behavior is most likely to occur. In addition, Touchette and his collea-
gues (Touchette, MacDonald, & Langer, 1985) described a method to assess the
time of day that the problem occurs most often by using a scatter plot. To create a
scatter plot, someone in the client’s natural environment records once each half
hour whether the problem behavior occurred during the preceding half hour. Scatter
plot recording is an interval recording method (Chapter 2), but is not an ABC obser-
vation method because antecedents and consequences are not observed and
recorded. After recording on the scatter plot for several days, you may be able to see

246 Chapter 13

the time of day that the problem behavior most often occurs. If the scatter plot shows
that the problem behavior usually occurs at certain times of the day, you can then
conduct ABC observations at those times. If the scatter plot does not reveal a pattern
in the time of occurrence of the problem behavior (e.g., see Kahng et al., 1998),
then ABC observations would need to be scheduled for longer periods or more time
periods in an attempt to observe the behavior. A scatter plot is shown in Figure 13-1.

Ti
m

e
of

D
ay

(3
0-

m
in

ut
e

in
te

rv
al

s)

9

10

11

12

1

2

3

4

5

6

7

8

9

10

11
2 4 6 8 10 12 14

Successive Days

Client Starting Date

FIGURE 13-1 This is a scatter plot recording sheet for documenting the time of day that the problem behavior
occurs. Each square on the grid represents a half-hour period in a particular day. To complete the
scatter plot, an observer records each half hour whether the problem occurred within the preceding
half hour. If the behavior occurred once in the half-hour time period, the observer puts a horizontal
slash through the box. If the behavior occurred two or more times, the observer darkens the box.
The observer leaves the box blank if the behavior did not occur in a particular half hour. After
recording on the scatter plot grid for a week or two, you will be able to determine the time that
the behavior occurs most frequently. In this completed scatter plot, notice that the behavior is
occurring most frequently in the afternoon around 3 P.M. With this information, the behavior analyst
would conduct ABC observations around 3 P.M. to record the antecedents and consequences of the
behavior. (From Touchette, P. E., MacDonald, R. F., & Langer, S. N. [1985]. A scatter plot for iden-
tifying stimulus control of problem behavior. Journal of Applied Behavior Analysis, 18, 343–351.
Copyright © 1985 University of Kansas Press. Reprinted by permission of the publisher.)

Understanding Problem Behaviors through Functional Assessment 247

The observer conducting the ABC assessment must be trained to observe and
record the antecedents and consequences correctly each time the problem behav-
ior occurs. This means that the observer must be able to discriminate each
instance of the problem behavior so that he or she can record the events that
immediately preceded and followed the behavior. The observer must be trained
to describe antecedent and consequent events objectively, in terms of the specific
behavior of other people and changes in physical stimuli in the environment.
The observer must record antecedents and consequences immediately as they
occur to reduce reliance on memory.

ABC observations can be conducted in three ways: the descriptive method,
the checklist method, and the interval recording method.

â–  In the descriptive method, the observer writes a brief description of the
behavior and of each antecedent and consequent event each time that the behav-
ior occurs. The observer typically uses a three-column data sheet similar to the
one shown in Figure 13-2. This method is open-ended and results in descriptions
of all events that were contiguous to the behavior. Because it is open-ended and
the observer describes all antecedent and consequence events that were observed,
this ABC assessment method may be conducted before indirect methods are used,
before any hypotheses are developed about the function of the behavior.

â–  The checklist method for conducting ABC observations involves a checklist
with columns for possible antecedents, behaviors, and consequences. The check-
list typically is developed after the problem behaviors and potential antecedents
and consequences are identified in an interview (or other indirect assessment

OBSERVATION RECORD

1 Describe the behavior(s)

2 Describe what happened just before the behavior occurred (what you did, what they did, etc.).

3 Describe what happened just after the behavior occurred (what you did, what they did, etc.).

Date,
time

What happened just
before the behavior?

Behavior: What was done
or said? Be specific.

What happened just
after the behavior?

FIGURE 13-2 This ABC observation data sheet includes columns to record the antecedents, the behavior, and the con-
sequence of the behavior. Each time the problem behavior occurs, the observer immediately writes
down a description of the antecedent events, the behavior, and the consequent events. With this ABC
observation method, the observer must be able to take the time to describe the events as they occur.

248 Chapter 13

method) or through observation. To conduct an ABC observation using the
checklist, the observer records the particular problem behavior each time it
occurs, together with its antecedents and consequences, by putting a check mark
in each of the relevant columns. Figure 13-3 shows an example of an ABC obser-
vation checklist.

â–  The interval (or real-time) recording method is the third way to conduct
ABC observations. Recall that in interval recording, you divide an observation
period into brief time intervals and mark a data sheet at the end of each interval
to record whether the behavior occurred in that interval, and in real-time record-
ing, you record the exact time of each occurrence of the behavior. You can also
identify and define specific events that may serve as antecedents and conse-
quences and record these events, as well as the behavior, with interval or real-
time recording. You identify the specific events to record from an interview or
other indirect assessment methods or through direct observation.

Rortvedt and Miltenberger (1994) conducted ABC observations using interval
recording to identify the function of noncompliant behavior in two young chil-
dren. Noncompliance was defined as refusing to complete a task requested by the
parent. The researchers first conducted an interview with the parents to assess the
function of the noncompliance. The parents of both children reported that they
responded to their child’s noncompliance with attention. They said that when
the child refused to carry out a requested activity, they repeated the request,
scolded the child, made threats of punishment, or pleaded with the child.
On the basis of this information, the researchers hypothesized that attention from
the parent was reinforcing noncompliance. The researchers conducted ABC

Behaviors Antecedents Consequences

Pe
rs

on

Ob
se

rv
at

io
n

Re
co

rd
in

g
Fo

rm

Ti
m

e

FIGURE 13-3 This ABC observation checklist includes a column for the time the behavior occurred and columns
for specific antecedents, behaviors, and consequences. Each time the behavior occurs, the observer
records the time and puts check marks in the columns indicating which behavior occurred, which
antecedent event occurred before the behavior, and which consequence followed the behavior. The
observer can record the ABCs quickly, without much disruption of ongoing activities. The target
behaviors, antecedents, and consequences are written at the top of the columns in each section
before recording. In this completed ABC observation checklist, the antecedent, behavior, and conse-
quence categories have been entered and some observations have been recorded.

Understanding Problem Behaviors through Functional Assessment 249

observations of the parent and child in the home. They asked the parent to make
a number of requests and then recorded the occurrence of child noncompliance
and parental attention after noncompliance using 10-second interval recording.
Their observations indicated that the children were noncompliant with 50%–80%
of the requests made by their parents. Furthermore, each time the child refused to
follow a request, the parent responded with attention. The results of the ABC
observations were consistent with the information from the interview and provided
strong support for the hypothesis that attention was reinforcing the noncompli-
ance. Successful treatment involved positive reinforcement for compliance and
a procedure called time-out (see Chapter 17), in which parental attention was
eliminated after noncompliance. This treatment was chosen on the basis of the
functional assessment results.

Together, indirect and direct functional assessment methods are categorized
as descriptive assessments because the antecedents and consequences are
described, either from memory or from direct observation of the events (Arndorfer
et al., 1994; Iwata, Vollmer, & Zarcone, 1990; Mace & Lalli, 1991; Sasso et al.,
1992). Descriptive functional assessments allow you to develop hypotheses about
the antecedent and consequent variables controlling the problem behavior, but
they do not prove that the variables are functionally related to the behavior.
To demonstrate a functional relationship, the antecedents or consequences must
be manipulated to show their influence on the problem behavior.

Direct Observation Assessment Methods
â–  Descriptive method

â–  Checklist method

â–  Interval or real-time method

Experimental Methods (Functional Analysis)
Experimental methods of conducting a functional assessment manipulate anteced-
ent or consequent variables to demonstrate their influence on the problem behav-
ior. Experimental methods are also called experimental analysis or functional
analysis. These terms reflect that these methods experimentally demonstrate a
functional relationship between the antecedents and consequences and the prob-
lem behavior. In a functional analysis, you follow the problem behavior with
potential reinforcers to see which consequences increase (strengthen) the behav-
ior, and/or you present different antecedent events (possible EOs) to see which
ones evoke the behavior.

Some researchers have manipulated both antecedents and consequences to
evaluate the possible functions of a problem behavior. For example, Iwata, Dor-
sey, Slifer, Bauman, and Richman (1982) conducted experiments to evaluate the
function of self-injurious behavior (SIB) exhibited by people with intellectual dis-
abilities. In the experimental conditions, Iwata arranged possible establishing
operations as an antecedent and possible reinforcing consequences for the SIB.
For example, to evaluate attention as a possible reinforcing consequence for SIB,
Iwata arranged a condition in which the child did not receive any attention from

250 Chapter 13

the adult who was present (EO), and then when SIB occurred, the adult provided
attention in the form of social disapproval. To evaluate escape from tasks as a pos-
sible reinforcer, Iwata presented difficult tasks (EO), and then when SIB occurred,
allowed the individual to briefly escape from the task. Iwata and colleagues evalu-
ated four conditions within an alternating-treatments design (Figure 13-4) and
showed that some of the children’s SIB was maintained by attention, others by
escape, and others by automatic reinforcement.

Other researchers have conducted functional analyses in which antecedents were
manipulated to determine their influence on the problem behavior. The function of
the problem behavior was then inferred from the resulting behavior changes associ-
ated with the antecedent manipulations. For example, Carr and Durand (1985)
conducted conditions involving decreased attention and increased task difficulty for
children with behavior disorders in the classroom. When problem behaviors were

•

•

•
• •

•

•

•

•

•

•

•

•

•

• • • • • • • • • •

•
• ••

•
•

Pe
rc

en
ta

ge
o

f 1
0-

Se
co

nd
In

te
rv

al
s

(S
IB

)

0

10 20 30

80

20

40

100

60

Child 1

0

10 20 30 40

80

20

40

100

60

Child 2

0

10 20 30

80

20

40

100

60

Child 4

0

10 20 30

80

20

40

100

60

Child 5

Academic
Alone
Social Disapproval
Play

Sessions

FIGURE 13-4 Graphed data for four representative subjects in an alternating-treatments design study. The level of the
child’s self-injurious behavior was recorded in each of four experimental conditions: academic demand,
social disapproval, alone, and unstructured play. (From Iwata, B. A., Dorsey, M. F., Slifer, K. J., Bauman,
K. E., & Richman, G. S. [1994]. Toward a functional analysis of self-injury. Journal of Applied
Behavior Analysis, 27, 205. Copyright © 1994 by Society for the Experimental Analysis of Behavior.
Reprinted by permission of the Society for the Experimental Analysis of Behavior.)

Understanding Problem Behaviors through Functional Assessment 251

greatest in the decreased attention condition, the authors inferred that the behavior
was maintained by attention because it was evoked in the presence of the EO for atten-
tion. When the problem behavior was greatest in the increased task difficulty condi-
tion, the authors inferred that the behaviors were maintained by escape from the task
because it was evoked by the EO for escape. Carr and Durand showed that some chil-
dren’s problem behaviors were greatest in the decreased attention condition, and
others were greatest in the increased task difficulty condition.

Sometimes functional analyses are designed to evaluate a range of possible
functions for the problem behavior (Iwata, Dorsey, et al., 1982). In such cases,
the behavior analyst may not have a hypothesis about the reinforcing consequence
maintaining the problem behavior and is exploring a range of possibilities in the
functional analysis. We would consider this form of a functional analysis an
exploratory functional analysis. An exploratory functional analysis typically
includes three or four test conditions and a control condition. In each test condi-
tion you present an EO and a possible reinforcer for the problem behavior; in a
control condition you present an AO and withhold the possible reinforcers. For
example, if you did not have clear hypotheses about the function of a problem
behavior, you might conduct four different conditions evaluating whether
attention, tangible reinforcers, escape, or sensory stimulation was the reinforcing
consequence for a problem behavior (Iwata, Dorsey, et al., 1982; Ellingson,
Miltenberger, Stricker, Garlinghouse, et al., 2000; Rapp, Miltenberger, Galensky,
Ellingson, & Long, et al., 1999). Exploratory functional analyses evaluating a
range of possible reinforcing consequences can identify a particular function of
a problem behavior while ruling out other functions.

In some cases, a functional analysis may involve fewer experimental condi-
tions because the behavior analyst is basing the conditions on a specific hypothesis
about the function of the problem behavior (Arndorfer et al., 1994). In such cases,
the goal of the functional analysis is not to evaluate all possible functions, but to
confirm or disconfirm the hypothesis. We would consider this form of a func-
tional analysis a hypothesis-testing functional analysis. In this type of functional
analysis, one condition (the test condition) presents the hypothesized EO, and
when the problem behavior occurs, presents the hypothesized reinforcer. The
other condition (the control condition) presents the hypothesized AO, and if the
problem behavior occurs, does not provide the hypothesized reinforcer. For exam-
ple, if you believed the target behavior was reinforced by attention, you might
evaluate two experimental conditions in a functional analysis: a test condition
involving no attention as an antecedent condition (EO) with attention contingent
on the target behavior and a control condition involving high levels of attention as
an antecedent condition (AO) with no attention after the target behavior. If the
target behavior occurred at a greater rate in the test condition and at a lower rate
in the control condition, the results would confirm the hypothesis that attention
was the reinforcing consequence for the target behavior.

How did Rich conduct a functional analysis of Jacob’s problem behavior in the
example presented earlier?

Rich manipulated the way the other children in day care interacted with Jacob.
Rich had developed a hypothesis that the antecedent for Jacob’s head-banging,
whining, and toy-throwing was that the other children touched or played with

252 Chapter 13

Jacob’s toys. To analyze whether this antecedent event was functionally related to
the problem behaviors, Rich arranged conditions in which this antecedent was pres-
ent (test condition) and conditions in which it was absent (control condition). Fur-
thermore, Rich hypothesized that the reinforcer maintaining Jacob’s problem
behaviors was the act of returning toys to him. To analyze whether this conse-
quence was functionally related to the problem behaviors, Rich arranged a condi-
tion in which the consequence was present (test condition) and a condition in
which it was absent (control condition). The results showed that in the test condi-
tion when the antecedent and consequence were present, the problem behaviors
occurred at a greater rate. When the antecedent and consequence were absent
(control condition), Jacob engaged in little head-banging, whining, and toy-
throwing behaviors (Figure 13-5). Thus, Rich demonstrated a functional relation-
ship between these particular antecedent and consequent events and the problem
behaviors for Jacob. The results supported the hypothesis that Rich had formed
based on the results of the interview and ABC observation assessments. The treat-
ment Rich implemented was successful because it was based on the results of the
functional assessment. That is, when Rich understood why Jacob was engaging in
the problem behaviors, he could develop an appropriate treatment.

In a similar manner, Rich conducted a functional analysis of Anna’s hitting,
kicking, and screaming. He had developed the hypothesis that the problem beha-
viors were most likely to happen when Anna’s mother was not paying attention to

Pe
rc

en
ta

ge
o

f I
nt

er
va

ls
o

f P
ro

bl
em

B
eh

av
io

rs

0

100

50

Anna

Jacob

•

•

•
•

•

•

•

HA LA HA LA FCT Rev FCT

•

•

•

•

•

•

•

UP IP UP IP FCT Rev FCT

Sessions

0

100

50

FIGURE 13-5 This graph shows the data from the functional analysis conducted with Anna and Jacob. For Anna, the
problem behavior increased in the low-attention (LA) condition and decreased in the high-attention (HA)
condition. This demonstrates that her mother’s attention was reinforcing the problem behavior. Func-
tional communication training (FCT) was the treatment procedure (see Chapter 15). Each time treatment
was implemented, the behavior decreased. For Jacob, UP is the uninterrupted-play condition and IP is
the interrupted play condition. His problem behavior was more frequent when other children interrupted
his play and returned his toys to him after he exhibited a problem behavior. This confirmed the hypothe-
sis that the return of his toys was the reinforcer for Jacob’s problem behaviors. When FCT was imple-
mented, the problem behavior decreased to low levels. For Anna and Jacob, Rev refers to reversal, a
condition in which functional communication was not used. After the reversal, FCT was implemented
again. (From Arndorfer, R. E., Miltenberger, R. G., Woster, S. H., Rortvedt, A. K., & Gaffaney, T.
[1994]. Home-based descriptive and experimental analysis of problem behaviors in children. Topics
in Early Childhood Special Education, 14, 64–87. Reprinted by permission of the author.)

Understanding Problem Behaviors through Functional Assessment 253

her, and that the reinforcer was her mother’s attention after she exhibited the
behaviors. Rich manipulated these antecedent and consequent events and found
that his hypothesis was supported. Furthermore, because the treatment based on
the results of the functional assessment was effective, it further supported those
results. The results of the functional analysis of Anna’s problem behaviors are
shown in Figure 13-5.

Types of Functional Analyses

Exploratory: A number of possible reinforcers are evaluated in the functional analysis (e.g., attention,
escape, tangible) along with a control condition in which no EOs or reinforcers for the problem behavior
are present.

Hypothesis testing: One test condition and one control condition are used to test the hypothesis that a
particular reinforcing consequence is maintaining the problem behavior.

Functional Analysis Research

There is substantial research on the use of functional analysis to identify the vari-
ables controlling problem behaviors in children and people with developmental
disabilities (Arndorfer & Miltenberger, 1993; Asmus et al., 2004; Hanley, Iwata, &
McCord, 2003; Iwata, Pace, et al., 1994; Kurtz et al., 2003; Lane, Umbreit, &
Beebe-Frankenberger, 1999; Mace, Lalli, Lalli, & Shea, 1993; Repp & Horner,
1999; Sprague & Horner, 1995). Carr, Newsom, and Binkoff (1980) conducted a
functional analysis of aggressive behavior in two boys with intellectual disabilities.
The researchers hypothesized that the antecedents to aggressive behavior were aca-
demic demands, and that escape from demands was the reinforcer for the problem
behaviors. To test this hypothesis, they arranged two experimental conditions: In
the first condition, academic demands were presented to the two children; in the
second condition, no demands were placed on the children. Carr found that
the aggressive behavior occurred at a high rate when demands were made, but
that the problem behavior decreased substantially when no demands were made
on the children. Because the children engaged in aggressive behavior in high-
demand conditions, it suggested that escape from demands was the reinforcer for
the aggressive behavior. Other research by Carr and Durand (1985) and Durand
and Carr (1987, 1991, 1992) has shown that the problem behaviors of students
with autism and intellectual disabilities may be reinforced by teacher attention or
by escape from the academic demands in the classroom. In each of these studies,
the researchers manipulated antecedent variables of teacher attention or task diffi-
culty to show a functional relationship between these variables and the problem
behaviors, and implemented effective treatments based on the function of the
problem behavior for each child. Figure 13-6 shows the functional analysis data
from Durand and Carr (1987).

Research by Iwata and his colleagues has illustrated the use of functional
analysis methods for identifying the controlling variables for SIB. Iwata, Dorsey,
et al. (1982) worked with children and adolescents with developmental disabilities
who were admitted to a hospital for the treatment of severe SIB. The researchers
arranged different experimental conditions to determine whether the reinforcer

254 Chapter 13

for the SIB was attention from adults, escape from demands, or the sensory stimu-
lation produced by the behavior itself. In the attention condition, an adult worked
on a task (ignoring the child) and paid attention to the child only after the occur-
rence of SIB. The attention from the adult involved statements of concern and
prompts to stop the behavior and to engage in toy play or other activities. This
condition was designed to simulate the common adult response to SIB. In the

Pe
rc

en
ta

ge
o

f I
nt

er
va

ls

0

25

0 10 3020

Sessions

•

•

• •

•
• •

•
•

•
• • • • • ••

• • •
•

•

•
• •

• • •

• •• •

• •

•

•

•
•

• •
•

•• •
• •• •

•
•

•
•

•

•

•
•

••

•• •• ••
•

• •

•

•

•

•

• •

•
•

•

•

• •• •• •••

•• •• •• •• •

•
•

•
•

• •

•

•

• •

• •
•

••
•

•
•

•
•0

50

0

15

0

50
Baseline Baseline

Jim

Decreased
Attention Baseline

Decreased
Attention

Increased
Task Difficulty Baseline Baseline

Increased
Task Difficulty

Baseline Baseline
Increased

Task Difficulty Baseline
Increased

Task Difficulty
Decreased
Attention Baseline Baseline

Decreased
Attention

Baseline Baseline
Increased

Task Difficulty Baseline
Increased

Task Difficulty
Decreased
Attention Baseline Baseline

Decreased
Attention

Baseline Baseline
Decreased
Attention Baseline

Decreased
Attention

Increased
Task Difficulty Baseline Baseline

Increased Task
Difficulty

Ro
ck

in
g

Ken

Ro
ck

in
g

Bob

H
an

d-
Fl

ap
pi

ng

Len

H
an

d-
Fl

ap
pi

ng

FIGURE 13-6 This graph shows the functional analysis of two problem behaviors, rocking and hand-flapping, in
four children with disabilities. Durand and Carr (1987) conducted three conditions: a baseline,
in which students worked on easy tasks and received substantial attention; a decreased attention
condition, in which they worked on an easy task but received much less teacher attention; and an
increased task difficulty condition, in which they received substantial attention but worked on more
difficult tasks. The graph shows that the problem behaviors were most frequent in the increased
task difficulty condition; this suggests that the problem behaviors were reinforced by escape from
the difficult tasks. (From Durand, V. M., & Carr, E. G. [1987]. Social influence of “self-stimulatory”
behavior: Analysis and treatment application. Journal of Applied Behavior Analysis, 20, 119–132.
Copyright © 1987 University of Kansas Press. Reprinted by permission of the publisher.)

Understanding Problem Behaviors through Functional Assessment 255

escape condition, an adult made typical educational demands on the child and,
after SIB, the adult terminated the demands for a brief period. This condition
was designed to simulate the situation that often occurs in the classroom when a
child engages in SIB. Finally, in the alone condition, the child was put into
a room alone without any toys or stimulating activities for a brief period.

Iwata compared the levels of SIB in the three experimental conditions.
If the rate of SIB was high in the attention condition and low in the other con-
ditions, it would demonstrate that attention was maintaining the SIB. If the level
of SIB was high only in the demand condition, it would demonstrate that SIB
was maintained by escape from demands. If the rate of SIB was high in the
alone condition, it would demonstrate that the SIB was maintained by the
sensory consequences produced by the behavior. Because the child was alone
without any adult interaction or stimulating activities, SIB in this condition
could not be reinforced by attention or escape and was presumed to be self-
stimulating. Iwata called this automatic reinforcement because the behavior
produces a reinforcing consequence automatically, without any response from
other people in the environment.

Iwata and colleagues demonstrated that the SIB of different children had dif-
ferent functions. For some children, the SIB was reinforced by attention, for
others by escape, and for some by sensory stimulation (automatic reinforcement).
This demonstration that SIB was maintained by different types of reinforcers with
different children was important. In later research, Iwata and his colleagues con-
ducted functional analyses of SIB exhibited by many other people with disabilities,
and they demonstrated effective treatments for the SIB in these people (Iwata,
Pace, Kalsher, Cowdery, & Cataldo, 1990; Iwata, Pace, Cowdery, & Miltenberger,
1994; Lerman & Iwata, 1993; Pace, Iwata, Cowdery, Andree, & McIntyre, 1993;
Smith, Iwata, Vollmer, & Zarcone, 1993; Vollmer, Iwata, Zarcone, Smith, &
Mazaleski, 1993; Zarcone, Iwata, Hughes, & Vollmer, 1993). Iwata’s findings, in
conjunction with the findings from Carr and Durand and others, suggest that you
must conduct a functional assessment of problem behaviors to understand their
functions and to choose the most effective treatments.

Investigators have continued to conduct research to refine functional assess-
ment methods and to establish their utility for choosing functional interventions
to address the factors contributing to the occurrence of problem behaviors. Recent
research has addressed a variety of problem behaviors exhibited by varied types of
individuals (e.g., McKerchar & Thompson, 2004; Moore & Edwards, 2003;
Ndoro, Hanley, Tiger, & Heal, 2006; Wallace & Knights, 2003; Wilder, Chen,
Atwell, Pritchard, & Weinstein, 2006).

There are advantages and disadvantages in using experimental methods (func-
tional analysis) for the functional assessment of problem behaviors. The primary
advantage is that a functional analysis demonstrates a functional relationship
between the controlling variables and the problem behavior. The functional anal-
ysis provides the standard of scientific evidence that a particular type of antecedent
evokes the behavior and a particular type of reinforcing consequence maintains
the behavior. Descriptive methods provide less certainty, although they do allow
us to formulate hypotheses about the controlling variables. The major disadvan-
tage of conducting a functional analysis is the time, effort, and professional

256 Chapter 13

expertise needed to manipulate the antecedents and consequences and measure
the resulting change in the behavior. A functional analysis is actually a brief exper-
iment, and people must be trained to carry out such an experiment. Most pub-
lished research on the functional assessment and treatment of problem behaviors
relies on functional analysis methods, whereas practitioners using behavior modifi-
cation procedures most often rely on descriptive functional assessment methods
(Arndorfer & Miltenberger, 1993).

FOR FURTHER READING
Research in Functional Analysis Methods

Once researchers demonstrated the utility of functional analysis procedures, they began investigating varia-
tions in functional analysis procedures for better understanding the reinforcing contingencies maintaining
problem behaviors. For example, some researchers have investigated the role of establishing operations in
functional analysis outcomes (e.g., Call, Wacker, Ringdahl, & Boelter, 2005; McComas, Thompson, &
Johnson, 2003; O’Reilly et al., 2006). Other researchers have evaluated the influence of session duration
(Wallace & Iwata, 1999) or the difference between brief versus extended functional analyses (Kahng &
Iwata, 1999) on functional analysis outcomes. Still other researchers have evaluated other influences,
such as instructions (Northup, Kodak, Grow, Lee, & Coyne, 2005) or medication (Dicesare, McCadam,
Toner, & Varell, 2005), on functional analysis outcomes. In another interesting investigation, researchers
evaluated the use of telemedicine to conduct functional analyses of children’s problem behavior in rural
settings. They showed that parents or teachers could conduct functional analyses when directed by
researchers via an interactive video network (Barretto, Wacker, Harding, Lee, & Berg, 2006).

Conducting a Functional Assessment

You should always conduct some form of functional assessment before you develop
treatment for a problem behavior. To develop the most appropriate treatment, you
should understand the environmental events (antecedents and consequences) that
control the behavior. Information on antecedents and consequences is important
because treatment will involve manipulating antecedents or consequences to pro-
duce a change in the behavior (see Chapters 14-16). You will need to know the
antecedents that evoke the problem behavior to use antecedent control procedures,
and you will need to know what the reinforcing consequence for the behavior is to
use extinction and differential reinforcement procedures effectively.

1. Start with a behavioral interview. Your functional assessment of a problem
behavior should start with an interview with the client or other informants
who know the client well and have specific knowledge of the problem
behaviors.

2. Develop a hypothesis about the ABCs of the problem behavior. The
outcome of the interview should be a clear definition of the problem beha-
viors and the development of hypotheses about the antecedents that evoke
the behaviors and the reinforcing consequences that maintain them. This
chapter focuses on this core information about controlling variables, but
the interview can also yield valuable information on alternative behaviors,
setting events or ecological variables, other reinforcing stimuli, and
previous treatments (see Table 13-2).

Understanding Problem Behaviors through Functional Assessment 257

3. Conduct a direct observation assessment. Once you have developed a
hypothesis about the controlling variables based on information from the
interview, the next step in the functional assessment is to conduct direct
observations of the ABCs in the natural context. ABC observations may be
conducted by the client, a professional consultant, or people in the client’s
environment who are trained by the consultant to conduct the observa-
tions. For example, a school psychologist might observe a child with a
problem behavior in the classroom, or the psychologist might train the
teacher or teacher’s aide to conduct the ABC observations. It is important
to take steps to reduce reactivity of the observations so that the information
on the ABCs reflects the typical level of the behavior and the typical ante-
cedents and consequences. Reactivity can be reduced through unobtrusive
observation, by participant observation, or by allowing a period of time
for the people in the natural setting to become accustomed to the
observer. If the information from the ABC observations is consistent with
the information from the interview, the initial hypothesis about antece-
dents and consequences is strengthened.

4. Confirm your initial hypothesis about the ABCs of the problem behavior.
With a firm hypothesis from multiple sources of assessment information
(interview and direct observation), you can develop functional treatments
that address the antecedents and consequences identified in the functional
assessment.

5. Conduct further assessments if needed. If the information from the ABC
observations is not consistent with the interview information, another inter-
view and further observations are needed to clarify the inconsistencies.
If further descriptive assessments produce consistent information that
allows you to develop firm hypotheses about the controlling antecedents
and consequences, you can consider the functional assessment complete
and develop functional interventions.

6. Conduct a functional analysis if needed. If the information from interviews
and ABC observations is still inconsistent after further assessment, a func-
tional analysis is necessary. A functional analysis is also warranted if the
information from the descriptive assessments is consistent but does not
lead to a firm hypothesis. Consider the following example.

Clyde, a young man with Down syndrome, started a job as part of a three-
person work crew that cleaned hotel rooms, with a job coach providing training
and supervision. When Clyde was asked to dust the dresser or the desk in a
room, he dropped to the floor, sat with his head down, and refused to work. The
job coach tried to talk Clyde into getting up and doing his job. She repeated the
request, explained to Clyde why he needed to work, and offered rewards, but
Clyde continued to sit on the floor. After a week in which the problem happened
on a daily basis, the job coach called a consultant for assistance. On the basis of
information from an interview with the job coach and from ABC observations,
the consultant found that Clyde engaged in the problem behavior every time he
was asked to work, and that the job coach consistently tried to talk him into work-
ing each time he refused.

258 Chapter 13

On the basis of this information, what are two possible hypotheses about the
reinforcer for the problem behavior?

One possibility is that attention from the job coach reinforced Clyde’s behav-
ior. A second possibility is that escape from the dusting task reinforced Clyde’s
problem behavior. The only way to determine which outcome reinforced the
problem behavior is to conduct a functional analysis in which these two possible
reinforcers are manipulated.

How would you conduct a functional analysis of Clyde’s behavior to identify the
reinforcer that is maintaining the behavior?

The two variables you want to manipulate are escape and attention as conse-
quences of the behavior. To manipulate these two variables, you arrange two condi-
tions: attention but no escape and escape but no attention. To arrange the first
condition, you tell the job coach to ask Clyde to dust, and when he drops to the floor,
to provide verbal prompts and physical guidance to prompt him to get up and dust the
table. In this condition, he is not escaping from the task (because the job coach is
using hand-over-hand guidance to get him to dust) but he is continuing to receive
attention contingent on refusing to work. In the second condition, you tell the job
coach to ask Clyde to dust, and when he drops to the floor, to provide no reaction. In
this condition, Clyde is escaping from the task but is not receiving attention. The job
coach arranges the two conditions on alternating days to see which condition pro-
duces the high rate of the problem behavior. If Clyde refuses to work more frequently
in the first condition, it suggests that the problem behavior is reinforced by attention.
If the rate of the behavior is higher in the second condition, escape is determined to
be the reinforcer for the problem. If the rate of the problem behavior is high in both
conditions, it suggests that the behavior is reinforced by both attention and escape.

The results of this functional analysis showed that Clyde refused to work most
in the second condition, thus suggesting that escape was the reinforcer for his
refusal to work. On the basis of these results, treatment was developed to address
the escape function of the behavior. Staff provided reinforcers for working (snacks
and brief breaks) and removed the reinforcer for refusing to work by manually
guiding him through the task each time he refused (see Chapters 14, 15, and 18
for details on these procedures).

As you can see from this example, a functional analysis does not have to be
complex or difficult to implement. The essential features of a functional analysis
are to have a reliable method of data collection to record the behavior in the dif-
ferent experimental conditions, to manipulate the antecedent or consequence
while holding other variables constant, and to repeat the experimental conditions
using a reversal design (or other experimental design) to demonstrate experimental
control over the behavior.

Functional Interventions

Once you have conducted the functional assessment process, you use the information
on antecedents and consequences of the problem behavior to develop interventions.
Interventions should be designed to alter the antecedents and consequences of the

Understanding Problem Behaviors through Functional Assessment 259

problem behavior to decrease the problem behavior and increase desirable alternative
behaviors. These functional interventions include extinction, differential reinforce-
ment, and antecedent manipulations. They are considered functional because they
address the antecedents and consequences identified in the functional assessment
(they address the function of the behavior). In addition, they are nonaversive because
they do not rely on punishment. Chapters 14, 15, and 16 describe these functional
interventions for problem behaviors.

CHAPTER SUMMARY

1. Conducting a functional assessment of a problem
behavior is the first step in developing a treatment
for the problem. The functional assessment helps
you to identify the antecedents that evoke the
behavior and the reinforcing consequence that
maintains the behavior.

2. A functional assessment may be conducted in three
ways: indirect assessment, direct observation assess-
ment, and experimental or functional analysis.

3. In an indirect assessment, you gather information on
the antecedents and consequences of the target
behavior from informants (people who know the cli-
ent well and are familiar with the problem behavior)
using behavioral interviews or questionnaires.

4. In a direct observation assessment (ABC record-
ing), you observe and record the antecedents,
behavior, and consequences as they occur in the
natural context. ABC recording can be done
using a descriptive method, a checklist method,
or an interval method.

5. Experimental methods for conducting a functional
assessment involve the manipulation of antecedents
or consequences to determine their influence on
the behavior. Experimental methods, also known
as functional analysis or experimental analysis,
allow you to establish a functional relationship
between the antecedents and consequences and
the problem behavior.

KEY TERMS

ABC observation, 246
descriptive assessments, 250
experimental analysis, 250

exploratory functional analysis, 252
functional analysis, 250
functional assessment, 237
functional interventions, 260

hypothesis-testing functional
analysis, 252

scatter plot, 246

PRACTICE TEST

1. What is a functional assessment of a problem
behavior? Why is it important to conduct a func-
tional assessment? (p. 240)

2. Identify and describe the four possible functions
of problem behaviors. (pp. 241–242)

3. Identify and describe the three major methods
for conducting a functional assessment of a
problem behavior. (pp. 242–254)

4. Identify and describe two ways to conduct an
indirect assessment. (pp. 243–245)

5. Identify a number of questions you could ask in an
interview to determine the antecedents and conse-
quences of a problem behavior. (pp. 244–245)

6. Identify and describe three ways to conduct ABC
direct observation assessments. (pp. 248–249)

7. What are descriptive functional assessment
methods? (p. 250)

8. Descriptive assessment methods do not de-
monstrate a functional relationship between
the antecedents and consequences and the
problem behavior. Explain that statement.
(p. 250)

9. What is the outcome of descriptive functional
assessment methods? (p. 250)

10. Describe how a functional analysis demon-
strates a functional relationship between the

260 Chapter 13

antecedents and consequences and the problem
behavior. (p. 250)

11. What is the difference between a functional
assessment and a functional analysis? (p. 242)

12. What is the first step in conducting a functional
assessment? (p. 257)

13. At what point would you consider your func-
tional assessment of a problem behavior com-
plete? Provide an example. (p. 258)

14. Under what circumstances do you need to
conduct a functional analysis of a problem
behavior? Provide an example. (p. 258)

15. Describe the three essential features of a func-
tional analysis. (p. 259)

16. Iwata and his colleagues found three types of
reinforcers for self-injurious behavior in children
and adolescents with developmental disabilities.
What are they? (p. 256)

17. Describe the three experimental conditions in
the functional analysis of self-injurious behav-
ior conducted by Iwata and his colleagues.
(pp. 254–256)

APPLICATIONS

1. If the goal of your self-management project is to
decrease an undesirable behavior, describe how
you will conduct a functional assessment of that
behavior. Describe each of the functional assess-
ment methods you will use to identify the con-
trolling variables for your target behavior.

2. Luther, an 80-year-old man, was recently
admitted to a nursing home because he had
Alzheimer’s disease and his wife could no longer
take care of him at home. Luther had spent his
life as a farmer. This was the first time he had
ever lived anywhere in which his freedom of
movement was restricted. Luther could not
leave the nursing home by himself and had to
learn to adapt to the daily routine in the nursing
home. Although the Alzheimer’s disease had
impaired Luther’s memory, he was still physi-
cally fit, and he enjoyed walking around the
building and talking to the staff and other resi-
dents. Shortly after he moved into the nursing
home, Luther started to exhibit a problem
behavior: He walked outside alone. He was not
allowed to go outside alone for safety reasons,
but he walked out the door a number of times
each day. When the weather was cold, he
walked outside without a coat. The staff had to
bring him back inside each time. The nursing
home has a main door near the nurse’s station,
another door near the business office, and three
fire doors at the sides and back of the building.
The building has four wings that form a square

with a totally enclosed courtyard in the center of
the building. Two doors open to the courtyard.
Four hallways, one down every wing of the
building, come together to form a square.
Assume that you are a behavioral consultant
who has been called by the nursing home staff
to help them deal with Luther’s problem behav-
ior. The staff does not know whether the prob-
lem is caused by the Alzheimer’s disease, which
causes Luther to become confused so that he
doesn’t know where he is or where he is going,
or whether it is the result of some contingencies
of reinforcement operating in the nursing home.
Your first step in developing a treatment strategy
is to conduct a functional assessment to deter-
mine why the problem is occurring. You have
scheduled a group interview with some of the
staff members who work regularly with Luther.
Provide a list of the questions you will ask the
staff to assess the antecedent events, the problem
behavior, and the consequences of the problem.

3. Interview questions from Application 2 and their
answers are provided here.

Problem behavior:

Q: What exactly does Luther do when he walks
outside?

A: He just walks up to the door, opens it, and
starts to go outside.

Q: Does he say or do anything as he is walking out
the door?

Understanding Problem Behaviors through Functional Assessment 261

A: He sometimes mumbles to himself about going
to see his wife or going to see somebody. Or he
says he has to go outside without giving a rea-
son. Sometimes he says nothing and just walks
outside. He usually looks at the nurse who is at
the nursing station as he walks out the door.

Q: What does he do once he is outside?
A: He’s not outside for more than a few seconds

because a staff person goes after him and brings
him back in. Usually, he gets outside and just
stands a few feet from the door. Often, he turns
and looks back in the building. Sometimes, the
nurse sees him going for the door and stops
him before he even gets outside.

Antecedents:

Q: What is Luther usually doing right before he
walks out the door?

A: Usually, he is walking around the hallways or
hanging around by the door.

Q: Is he usually by himself or with somebody
when he walks out the door?

A: He likes to talk to people when he’s walking
around the halls, but he is most often alone
when he goes for the door.

Q: Which door is he most likely to go out?
A: He’s tried them all, but most of the time he

goes out by the main nurse’s station.
Q: Does he ever walk out the door to the

courtyard?
A: No, hardly ever.
Q: What time of day is he most likely to walk out?
A: Usually when the staff are the most busy: when

they are providing care routines with the other
residents, before meals when they are helping
other residents, and at shift changes.

Q: Is there someone at the nurse’s station when he
walks out?

A: Almost always. We have someone at the nurse’s
station almost all the time.

Q: Even at busy times?
A: Yes, usually the nurse is charting or doing

paperwork at the nurse’s station at that time.

Consequences:

Q: What happens as soon as Luther walks out the
door?

A: A staff person runs out after him and brings
him back. Usually it is the nurse or nursing
assistant at the nurse’s station who sees him
leave.

Q: What happens then?
A: The nurse or nursing assistant walks back with

Luther and tells him why he can’t go outside
by himself. The staff person usually takes him
to the break room and sits down with him for a
few minutes with a cookie or cup of coffee.
The staff tries to get him interested in some-
thing other than leaving. It usually takes
5 minutes or more each time he tries to go
outside.

Q: What would happen if Luther went out the
door to the courtyard?

A: He has done that only once or twice. When he
went out in the courtyard, staff left him alone
because it is enclosed and he couldn’t wander
away or harm himself. He doesn’t go out that
door anymore.

On the basis of this information, what is your
initial hypothesis about the function of Luther’s
problem behavior? Describe the ABC observation
procedure you will develop in conjunction with the
nursing home staff. Describe the data sheet you will
use and the instructions you will give to the staff to
carry out the direct observation procedure.

4. The ABC observation procedure for Luther is
described here, together with the information
derived from the ABC observation. Because
Luther almost always goes out the door by the
nurse’s station, the data sheet will be kept at the
nurse’s station. Having already gathered infor-
mation on probable antecedents and conse-
quences, the consultant will have staff record
ABCs using a checklist. The checklist will item-
ize the probable antecedents and consequences;
the staff will put a check mark in the column
that corresponds to the relevant events. The staff
will also record the time of the behavior. The
data sheet will have a column for the time of
the behavior, a column where the staff member
who observes the behavior puts his or her initi-
als, and columns for each of the antecedents
and consequences as follows.

262 Chapter 13

Antecedents:
â–  Luther is alone or no one is talking to him.
â–  Luther is walking the hallways.
■ Luther looks at the nurse at the nurse’s station as
he goes for the door or goes out the door.

Consequences:
â–  Staff run after Luther and walk him back.
â–  Staff talk to Luther as they walk with him.
â–  Staff spend time with Luther after he is back in the
building.

â–  Luther gets coffee or cookies.

Staff will record these events immediately each
time the problem behavior occurs for 1 week.

The results of the ABC observations were as fol-
lows. Luther exhibited the problem behavior an aver-
age of five times per day. Luther was alone or no one

was talking to him 100% of the times that the prob-
lem occurred. He was walking the hallways or hang-
ing around the door 100% of the time, and he
looked at the nurse in the nurse’s station 90% of
the time before he walked out the door. When
Luther walked out the door, 100% of the time a
staff person ran after him and talked to him as he
or she brought him back. A staff person spent a few
minutes with him every time but one, and he got
coffee and cookies 50% of the time.

Does this information support your initial hypoth-
esis developed from the interview? Explain. On the
basis of the information from the interview and ABC
observations, describe the functional analysis proce-
dure you will use to confirm your hypothesis about
the function of the problem. Describe the two func-
tional analysis conditions you will have the nurses
conduct with Luther. Describe the type of results
you expect from the functional analysis procedure.

MISAPPLICATIONS

1. Hanna, a first grade student, was exhibiting disrup-
tive behaviors in the classroom. She was out of her
seat frequently; she talked, teased other students,
and got into the supply cabinet. To decrease this
behavior, the teacher came up with the following
plan. He decided to ignore Hanna’s disruptive
behavior and to praise Hanna whenever she was
in her seat paying attention and not exhibiting dis-
ruptive behavior. He believed that the use of differ-
ential reinforcement (extinction of the disruptive
behavior and reinforcement of appropriate behav-
ior) would decrease the disruptive behavior while
increasing the appropriate behavior. What is
wrong with this plan?

2. After talking to the school psychologist, Hanna’s
teacher learned that before you decide on a
treatment for a problem behavior, you must con-
duct a functional assessment of the problem to
identify the environmental variables that are
causing the behavior. The school psychologist
wanted the teacher to collect information on
the antecedents and consequences of the prob-
lem behavior by conducting ABC observations
in the classroom. The psychologist gave the

teacher a data sheet with three columns: one
for antecedents, one for the problem behavior,
and one for the consequences. The psychologist
asked the teacher to keep the data sheet on his
desk and, each time Hanna exhibited a problem
behavior, to get the data sheet and write down a
description of the antecedents, a description of
the behavior, and a description of the conse-
quences. The psychologist told the teacher that
they could get a good understanding of why the
problem behavior was occurring if the teacher
would do this ABC recording each day for a
week. What is wrong with the functional assess-
ment method used in this situation?

3. The director of a residential program for people
with severe intellectual disabilities asked the staff
to do behavioral observations of two residents who
were having behavioral difficulties and to develop
hypotheses about why their behavior problems
were occurring. One resident, Robyn, engaged in
aggressive behavior in which she screamed at and
slapped staff when they asked that she engage in
some training activities. The other resident,
Melvin, engaged in disruptive behavior in which

Understanding Problem Behaviors through Functional Assessment 263

he knocked items off the table and grabbed recrea-
tional items (e.g., games, magazines, and needle-
work) from other residents. The staff hypothesized
that Robyn was frustrated with the daily expecta-
tions that staff had for her and the demands that
they placed on her. They hypothesized that she
was communicating her feelings of displeasure
with staff. For Melvin, staff hypothesized that he

was bored and jealous of other residents who were
engaging in recreational activities. They hypothe-
sized that his disruptive behavior was a demonstra-
tion of his boredom and jealousy. What is wrong
with this approach to functional assessment to iden-
tify the variables responsible for Robyn’s and
Melvin’s problem behaviors? How could the staff
improve their functional assessment?

264 Chapter 13

Chapter Fourteen

Applying Extinction

After you have conducted a functional assessment of a problem behavior, youwill implement functional treatment procedures to alter the antecedents and
consequences of the problem behavior. This chapter describes the use of extinc-
tion, one of the functional treatments used to eliminate a problem behavior. As
explained in Chapter 5, extinction is a basic principle of behavior in which elimi-
nating the reinforcing consequence for a behavior results in a decrease in the fre-

quency of the behavior. To use extinction, you must first identify
the reinforcer that maintains the problem behavior and then elimi-
nate it. A behavior that is no longer reinforced will decrease in fre-
quency and stop. Consider the following example.

The Case of Willy

Willy, a 54-year-old man with mild intellectual disability, recently
moved into a group home because his parents were no longer able to
take care of him. He had lived his entire life with his mother and father
before coming to the group home. In the group home, Willy exhibited
a problem behavior: He argued when he was asked by a staff person to
do a training activity such as cooking, cleaning, laundry, or another
independent living skill. The functional assessment interview and
ABC observations produced the following information on the behavior

problem, antecedents, and consequences. The antecedent situation was that a female
staff person asked Willy to perform a daily living task. Willy did not exhibit the prob-
lem behavior when a male staff person asked him to perform a task. The problem
behavior was that Willy verbally refused to do the task and made statements such as,
“That’s women’s work,” “A woman ought to do that,” or “That ain’t man’s work.” This
behavior continued for up to 15 minutes, but usually he completed the task
eventually.

The consequence of Willy’s behavior was that the female staff member
argued with Willy, told him that he was making sexist comments, and tried to
convince him that men have to do these tasks also. The female staff member

â–  Why is it important to conduct a
functional assessment before using an
extinction procedure?

â–  What five questions must you address
before using an extinction procedure?

â–  How does the schedule of reinforce-
ment for a behavior influence extinction?

â–  Why is it important to reinforce
alternative behaviors when using
extinction?

â–  How can you promote generalization
and maintenance after the use of
extinction?

265

often became visibly upset at Willy’s sexist remarks and usually argued with him
until he started to perform the task.

The assessment information led to the hypothesis that the antecedent event
was that a female staff member made a request for Willy to complete a task and
that the staff person’s attention (i.e., arguing, explaining, emotional reactions)
after the problem behavior was the reinforcing consequence. Negative reinforce-
ment (escape) did not appear to play a role because Willy eventually completed
the requested task.

The staff wanted to decrease the frequency of Willy’s sexist comments and
refusal to complete tasks. The functional assessment results suggested that to
decrease the problem behavior, female staff members would have to eliminate
their attention after the behavior. The group home manager held a meeting with
the staff to teach them how to use extinction with Willy.

First, she told the staff about the finding of the functional assessment: that female
staff members’ attention appeared to be reinforcing the problem behavior. She then
told them that they would have to eliminate the reinforcer for the problem behavior
for the behavior to decrease. She gave the staff the following instructions: “Whenever
you make a request for Willy to complete a task and he refuses or makes sexist com-
ments, do not repeat the request and do not respond to Willy in any way. Do not
argue with him. Do not try to talk him into doing the task. Do not try to explain to
him that his sexist remarks are unacceptable. Do not show Willy any kind of emo-
tional reaction. Do not make a face that looks disappointed or upset. Simply walk
away and engage in another activity when Willy engages in the problem behavior.”

After providing these instructions for the use of extinction, the group home
manager modeled the use of extinction for her staff. She had another staff person
role-play Willy refusing a request and making sexist comments and, in response,
she simply walked away and made no response to Willy’s problem behavior.
Next, she role-played Willy and had each of the staff members rehearse the use
of the extinction procedure in response to Willy’s problem behavior. After each
staff member had demonstrated the use of extinction in the role-play demonstra-
tions with different variations of Willy’s problem behavior, she instructed the staff
to use the procedure with Willy whenever he engaged in the problem behavior in
response to a request. She warned the staff that they all had to use the extinction
procedure consistently, and that they had to ignore Willy’s sexist remarks no mat-
ter how upsetting they were. She emphasized that if just one person continued
to respond to Willy’s problem behavior with attention, Willy would continue to
engage in the problem and the extinction procedure would not be successful.
She also warned the staff members that Willy might escalate in his behavior
when they started using extinction. His refusals might become louder or longer,

Antecedent Behavior

Outcome: Willy is more likely to refuse to do tasks and to make sexist remarks when female staff members make a request.

Female staff member makes a request. Willy refuses to complete the task,
makes sexist comments.

Staff provides attention (arguing, explaining).

Consequence

266 Chapter 14

and he might make more upsetting comments. The staff should be ready for this
extinction burst and continue to ignore this behavior.

In conjunction with this extinction procedure, the group home manager
instructed staff members to praise Willy as soon as he started to engage in the
task that they requested. She told the staff that they must reinforce Willy’s cooper-
ative behavior with their attention so that this behavior would increase as his prob-
lem behavior decreased. Because Willy would no longer receive staff attention for
refusing and making sexist comments, it was important for Willy to receive staff
attention for the desirable behavior.

To promote generalization of the behavior change, the group home manager
emphasized that all staff must use the extinction procedure (and the reinforce-
ment procedure) at all times and in all situations with Willy. This meant that all
new staff and substitute staff must be trained to use the procedure. Furthermore,
she had a meeting with Willy’s parents and asked for their help when Willy visited
them on the weekend. Because she did not want the behavior reinforced on the
weekends, she asked Willy’s parents to do one of two things. They could refrain
from asking Willy to do any tasks when he was home, or they could use the
extinction procedure in the same way that the staff was using it. By not asking
Willy to do any tasks, they would be using a stimulus control procedure in which
they removed the antecedent for the behavior problem so that the behavior prob-
lem would not occur. Willy couldn’t refuse to do a task if he was never asked to
do one. Because Willy’s mother had always done everything for him in the past
anyway, she was most comfortable with this option.

Staff members collected data on the percentage of times that Willy refused to
complete tasks and found that his refusals decreased over time once the extinction
procedure was implemented. He continued to refuse once in a while, but staff
did not reinforce the behavior and the refusals did not last very long. Most often,
Willy completed the tasks that staff had requested as soon as they asked him.

Antecedent Behavior

Outcome: Willy is less likely to refuse requests and make sexist comments in the future.

Female staff member makes a request. Willy refuses to complete the task,
makes sexist comments.

Staff member walks away, pays no attention.

Consequence

Antecedent Behavior

Outcome: Willy is more likely to comply with staff requests in the future.

Female staff member makes a request. Willy complies with the request. Staff member provides praise.

Consequence

Reinforcement

Extinction

Applying Extinction 267

This example illustrates the steps involved in using extinction to decrease a
problem behavior (Table 14-1).

Using Extinction to Decrease
a Problem Behavior

Extinction is one of the first approaches that should be considered for treating a
problem behavior. As long as a problem behavior continues, there must be a rein-
forcing consequence contingent on the behavior that is maintaining it. Therefore,
to decrease the behavior, you must identify the reinforcing consequence and elim-
inate it (whenever possible). When the problem behavior is no longer reinforced,
it will extinguish. Let’s examine the steps involved in using extinction procedures
effectively (Ducharme & Van Houten, 1994).

Collecting Data to Assess Treatment Effects
As discussed in Chapters 2 and 3, observation and recording of the target behav-
ior are important components of a behavior modification program. You must
record the problem behavior before and after the use of the extinction procedure
to determine whether the behavior decreased when extinction was implemented.
You will need a behavioral definition of the problem behavior to be decreased, a
reliable data collection method, a baseline assessment to determine the level of
the problem behavior before the use of extinction, data collection in all relevant
settings after treatment to determine whether the behavior decreased and
whether generalization occurred, and continued data collection over time to
assess the maintenance of behavior change. If you are conducting research to
experimentally evaluate the effects of the extinction procedure, an acceptable
research design (see Chapter 3) and assessment of observer reliability also are
needed. The basic point to remember is that if you are going to use an extinc-
tion procedure (or any other behavior modification procedure), you must collect
data on the problem behavior to document the change in behavior after the
use of the procedure. If your recording of the problem behavior shows that

TABLE 14-1 Step in Using Extinction

1. Collect data to assess treatment effects.

2. Identify the reinforcer for the problem behavior through functional assessment.

3. Eliminate the reinforcer after each instance of the problem behavior.

â–  Have you identified the reinforcer?

â–  Can you eliminate the reinforcer?

â–  Is extinction safe to use?

â–  Can an extinction burst (escalation of the problem behavior) be tolerated?

â–  Can consistency be maintained?

4. Consider the schedule of reinforcement for the problem behavior.

5. Reinforce alternative behaviors.

6. Promote generalization and maintenance.

268 Chapter 14

the behavior did not change after treatment, you can reassess the problem or the
implementation of the extinction procedure and make whatever changes are
necessary to decrease the problem behavior.

Identifying the Reinforcer for the Problem Behavior through
Functional Assessment
In functional assessment, you identify the antecedents and consequences of the
problem behavior (see Chapter 13). This is a critical step in using extinction proce-
dures effectively. You must identify the specific reinforcer for the problem behavior
so that you can eliminate it in an extinction procedure. You cannot assume that a
particular reinforcer is maintaining a problem behavior. The same problem behav-
ior exhibited by different people may be maintained by different reinforcers. For
example, one child’s aggressive behavior might be reinforced by the parents’ atten-
tion, whereas another child’s aggressive behavior might be reinforced by getting
toys from siblings. Sometimes, the same behavior exhibited by a particular person
in different situations might be maintained by different reinforcers (e.g., Romaniuk
et al., 2002). For example, a young child cries when she has trouble tying her
shoes, and the crying is reinforced when the parents help her tie the shoes. This
same child might cry when the parents make a request (e.g., to brush her teeth),
and the crying is reinforced when the parents allow her to escape from the task
that was requested. A behavior may serve different functions in different contexts
(Day, Horner, & O’Neill, 1994; Haring & Kennedy, 1990).

The success of an extinction procedure depends on whether the particular
reinforcer maintaining the problem behavior has been identified. A variety of
stimuli or events may function as reinforcers for problem behaviors. Problem
behaviors may be maintained by positive reinforcement when the behavior results
in the presentation of a stimulus or event or negative reinforcement when the
behavior results in escape from some stimulus or event. The reinforcing conse-
quence may involve the behavior of another person or a change in a physical
(nonsocial) stimulus or environmental event. Table 14-2 describes a variety of
problem behaviors and the consequences reinforcing those behaviors.

For each problem behavior in Table 14-2, identify whether the example illustrates
social positive reinforcement, social negative reinforcement, automatic positive rein-
forcement, or automatic negative reinforcement. (Answers appear in Appendix A.)

Eliminating the Reinforcer after Each Instance of the Problem
Behavior
Extinction, by definition, involves eliminating the reinforcer after each instance
of the problem behavior. Although this may seem straightforward, a number of
considerations must be addressed for the successful use of extinction.

Have You Identified the Reinforcer? Obviously, you cannot eliminate the rein-
forcer for the problem behavior until you have identified it by means of a func-
tional assessment. Failure to eliminate the particular stimulus or event that
functions as the reinforcer for the problem behavior is failure to implement the

Applying Extinction 269

extinction procedure correctly (Iwata, Pace, Cowdery, & Miltenberger, 1994;
Mazaleski, Iwata, Vollmer, Zarcone, & Smith, 1993).

The extinction procedure may be different depending on the reinforcer that is
maintaining the problem behavior. For example, when Iwata and his colleagues
worked with three children with developmental disabilities who engaged in self-
injurious behavior (SIB)—head-banging—they found that the reinforcer for the
SIB was different for each child (Iwata, Pace, Cowdery, & Miltenberger, 1994).
For one child, SIB was reinforced by attention from adults. For another child,
SIB was reinforced by escape from educational demands. For the third child, SIB
was reinforced automatically by the sensory consequences of the behavior itself.
Iwata demonstrated that the extinction procedure was different for each child
because the reinforcer for the SIB was different for each child.

How did Iwata implement extinction for the SIB that was reinforced by atten-
tion from adults?

Because the SIB was maintained by attention, extinction involved eliminating
the attention after each instance of SIB. This particular child, Millie, was an
8-year-old who banged her head on flat surfaces such as a wall or the floor. When
she banged her head, the adult who was present did not respond in any way, no
matter how long the head-banging continued (Figure 14-1). (It is important to
note that precautions were taken so that the child could not hurt herself.)

TABLE 14-2 Examples for Self-Assessment (Problem Behaviors and Reinforcers)

Problem Behavior Reinforcing Consequence

1. A child complains of being sick when told to
do chores.

A parent does the household tasks for the child.

2. A person with intellectual disability runs into
the street and refuses to leave the street.

A staff member offers a can of soda if the person
leaves the street.

3. A spouse has a temper outburst during a
disagreement.

The other spouse stops arguing and agrees to the
spouse’s demand.

4. A child with autism flicks his fingers in front
of his eyes.

This behavior produces visual stimulation.

5. A person runs away from a dog while walking
down the street.

The person gets away from the dog, and the fear
reaction diminishes.

6. A child refuses to comply with a parent’s
request to do a task.

The child avoids the task and continues to watch
television.

7. A child refuses to comply with a parent’s
request to do a task.

The parent repeats the request, pleads with the
child, and scolds the child.

8. A hospital patient calls the nurses’ station
several times a day.

A nurse comes to the room each time to check on
the patient but finds no problem.

9. A patient with a brain injury strips naked each
time the nurse enters the room for the
morning routine.

The nurse reacts with surprise and indignation and
orders the patient to get dressed.

10. A factory worker on an assembly line sabo-
tages the line so that it stops.

The factory worker sits down and has a cigarette
and a cup of coffee each time the line is down.

270 Chapter 14

However, the adult did provide attention when Millie did not bang her head. This
procedure, in which a reinforcer is delivered for the absence of the problem
behavior, is described in Chapter 15.

How did Iwata implement extinction for the SIB that was reinforced by escape
from educational demands?

Because the SIB was reinforced by escape from educational demands, extinc-
tion involved eliminating escape after the SIB. This particular child, Jack, was a
12-year-old boy. He engaged in the SIB in teaching situations when the teacher
asked him to perform learning tasks such as object identification or simple motor
tasks and his SIB resulted in getting out of doing the task (escape). To use extinc-
tion, when the SIB occurred, the teacher used physical guidance so that Jack
could not escape from the task. Regardless of how long he engaged in the SIB,
the teacher continued to present the educational demands and used physical guid-
ance to prevent escape. The teacher also provided praise when the child complied
with the educational tasks.

Donnie, the third child in Iwata’s study, was a 7-year-old boy who engaged
in SIB that was automatically reinforced. Because there was no social reinforce-
ment for the head-banging, it was presumed to be reinforced by the sensory

FIGURE 14-1 The child’s head-banging is reinforced by the adult’s attention. Notice that the child wears a helmet for
safety. The adult removes the reinforcer for head-banging by not paying attention to the behavior.
Because the behavior is no longer reinforced, it will stop.

Applying Extinction 271

consequences generated by the behavior. Iwata and his colleagues used sensory
extinction: They put a padded helmet on Donnie so that the sensory conse-
quences of the head-banging would be altered. If the head-banging no longer
produced the same sensory reinforcement, the behavior would extinguish. The
results showed that the SIB decreased when Donnie wore the padded helmet.

This study (Iwata, Pace, Cowdery, & Miltenberger, 1994) convincingly illus-
trates that to use extinction, you must identify the reinforcer for the particular
problem behavior and eliminate that reinforcer. If you do not identify the rein-
forcer for a particular problem behavior, you cannot use extinction. For example,
imagine a parent whose 3-year-old child gets cookies from the cookie jar fre-
quently throughout the day. The parent wants the child to stop taking cookies
from the cookie jar. Because of a limited understanding of extinction, the parent
ignores the behavior each time it occurs and believes that not providing attention
will decrease the child’s behavior. What is wrong with the parent’s action? The
problem is that taking cookies from the cookie jar is reinforced by eating cookies,
not by the parent’s attention. Therefore, eliminating the parent’s attention after
the behavior does not eliminate the reinforcer for the behavior. The behavior con-
tinues to be reinforced and, therefore, continues to occur (Martin & Pear, 1992).

How would the parent implement extinction in this case?

The parent would implement extinction by eliminating the reinforcer (the
cookies) for the problem behavior. If the parent took the cookies out of the cookie
jar, the problem behavior of going into the cookie jar would no longer be reinforced
by getting cookies. As a result, the child would quit going into the cookie jar.

For each problem behavior in Table 14-2, describe how you would implement
extinction. (Answers are listed in Appendix B.)

Functional Variations of Extinction
â–  Extinction following positive reinforcement. When a behavior is positively reinforced, extinction
means the person no longer gets the positive reinforcer following the behavior.

â–  Extinction following negative reinforcement. When a behavior is negatively reinforced,
extinction means the person no longer escapes from the aversive stimulus following the behavior.
This variation of extinction is called escape extinction.

Can You Eliminate the Reinforcer? After you have conducted a functional
assessment to identify the reinforcer for the problem behavior, you must deter-
mine whether the change agent (parent, teacher, staff member, nurse, client) can
control the reinforcer. If the change agent has no control over the reinforcer,
extinction cannot be implemented. For example, in the case of Willy’s noncom-
pliance and sexist comments, the reinforcer for Willy’s problem behavior was
attention from the staff. This reinforcer is under the control of the change agents,
the staff members. They can withhold their attention after the problem behavior,
and they can provide their attention after Willy’s cooperative behavior. Therefore,
they can successfully implement the extinction procedure.

For some problem behaviors, however, the change agents do not have control
over the reinforcer. If a grade school boy threatens to hurt other children to get

272 Chapter 14

their lunch money, the reinforcer for this behavior is the receipt of the money
(and perhaps other reactions of the victims). The teacher does not have control
over this reinforcer because the problem behavior happens when the teacher or
another adult is not present. Therefore, the teacher cannot use extinction. The
teacher could instruct the class not to give away their lunch money when they
are threatened, but it is likely that the problem behavior will still be reinforced at
least occasionally by children who continue to give away their money when they
are threatened. Consider another example.

A teenager plays her stereo so loudly that it disturbs the rest of the family.
The reinforcer for this behavior is the loud music. (Assume you have ruled out
the parents’ attention as the reinforcer.) Unless the parents have installed an elec-
tronic device on the stereo that does not permit the volume to be turned up
beyond a certain level, the parents do not have control over this reinforcer. The
teenager’s behavior of turning the knob on the stereo is immediately reinforced
each time by an increase in the loudness of the music. The parents might ask her
to turn it down or implement a punishment procedure to decrease the behavior,
but they cannot use extinction because the loudness of the music (the reinforcer)
is not under their control.

When considering the use of extinction to decrease a problem behavior, you
must determine that the change agent can control the reinforcer maintaining the
problem behavior. Extinction can be implemented only if the change agent can
prevent the reinforcing consequence each time the problem behavior occurs.

Is Extinction Safe to Use? Before deciding to use extinction, it is important to
determine whether extinction could result in harm to the person exhibiting the
problem behavior or to other people in the immediate environment. Consider
the following examples.

Rupert is a young man with severe intellectual disability who works in a shel-
tered workshop during the day. He sits at a table with three other people and
assembles parts for a local factory. Rupert engages in a problem behavior in
which he attacks the people at his table. He grabs people by the hair and hits
their heads on the table. When this happens, staff members immediately intervene
and separate Rupert from the other person. The functional assessment identified
staff attention as the reinforcer maintaining this problem behavior. Extinction
would require the staff to provide no attention after each instance of the problem.
However, it would be extremely harmful to the person being attacked if the staff
did not intervene immediately. In this case, therefore, extinction is not a safe
procedure and cannot be used.

Now consider the case of 4-year-old Annie, who runs out in the street when
she is playing in the front yard. The babysitter, who is usually sitting in the front
yard reading a book or magazine, yells for Annie to get out of the street. When
Annie refuses, the babysitter runs out into the street to get her. The reinforcer for
this behavior is the babysitter’s attention. However, extinction cannot be used in
this case because it is not safe to ignore a child when she runs into the street.
Other procedures, such as differential reinforcement or antecedent control, should
be used instead (see Chapters 15–18).

Ben is an 18-year-old man with intellectual disability who participates in a
residential training program. The staff is trying to teach him some basic self-care

Applying Extinction 273

skills, such as shaving and toothbrushing. The problem is that Ben engages in
aggressive behavior (hair pulling, scratching, and pinching) when a staff member
attempts to teach him these skills. When Ben grabs the staff member’s hair or
scratches or pinches the staff member, the session is terminated. As a result, Ben’s
aggressive behavior is negatively reinforced by escape from the training session.
Extinction in this case would involve continuing the training session when Ben
engaged in aggressive behavior so that the problem behavior did not result in
escape. However, it is dangerous for staff members to continue the session when
Ben is aggressive toward them, so it is difficult to use extinction. In this case, a
procedure such as response blocking or brief restraint might facilitate the use of
extinction (see Chapter 18).

As you now see, even if you have identified the reinforcer for the problem
behavior and the change agent has control over the reinforcer, you cannot use
extinction until you are certain that it is safe to eliminate the reinforcer. Extinc-
tion may be particularly unsafe when a problem behavior is negatively reinforced
because extinction requires that you prevent escape when the problem behavior
occurs. To prevent escape often requires physical guidance through the task,
which may be difficult or impossible if you are working with an adult who physi-
cally resists the guidance. In such cases other functional procedures (antecedent
manipulations, differential reinforcement) should be used instead of extinction.

Can an Extinction Burst (Escalation of the Problem Behavior) Be Tolerated?
As discussed in Chapter 5, the use of extinction often is accompanied by an
extinction burst, in which the behavior increases in frequency, duration, or inten-
sity, or novel behaviors or emotional responses occur (Goh & Iwata, 1994;
Lerman, Iwata, & Wallace, 1999; Vollmer et al., 1998). Before you decide to use
extinction, you must anticipate the extinction burst and be certain that the
change agents can tolerate the escalation in the behavior. Consider the case of a
5-year-old girl with bedtime tantrums. When taken to bed, she screams and cries.
After her parents leave the room, she calls for them. When she exhibits these
behaviors, the parents go into her room to calm her down and talk to her until
she is asleep. Their attention is reinforcing the problem behavior.

The parents could use extinction to decrease and eliminate the problem
behavior, but they must realize that as soon as they no longer respond to the tan-
trums, the child is likely to exhibit an extinction burst in which the problem
behavior escalates; she will engage in tantrum behavior that is more intense and
lasts longer. If the parents are not prepared for this outcome, the use of extinction
may fail. The first time the parents ignore the tantrums at bedtime and the prob-
lem behavior escalates, they might become concerned or frustrated and go into
the child’s room, thereby reinforcing the tantrum behavior. This is likely to make
the problem even worse because the parents will reinforce a worse instance of the
problem behavior (e. g., more intense, longer duration). As detailed in Chapter 9,
serious problem behaviors often are shaped in this way.

When using an extinction procedure, you must inform the change agent of
the escalation that is likely to occur during an extinction burst. Furthermore, you
must instruct the change agent to persist in withholding the reinforcer as the prob-
lem behavior escalates. If escalation of the behavior is likely to harm the person

274 Chapter 14

with the problem behavior or other people, you must devise a plan to eliminate or
minimize the harm. Iwata put a helmet on the young girl who engaged in head-
banging so that she would not cause harm to herself during the extinction proce-
dure (Iwata, Pace, Cowdery, & Miltenberger, 1994). Carr had teachers wear
protective clothing to protect them from the aggressive behavior of two boys
during an extinction procedure (Carr, Newsom, & Binkoff, 1980). You might
instruct parents to remove breakable objects from the room when using extinction
for their child’s disruptive behavior or tantrums to prevent damage to objects or
harm to the child.

If you predict that the change agent will be unable to persist in withholding
the reinforcer during an extinction burst, or if you cannot prevent harm during
an extinction burst, an extinction procedure should not be used. Other procedures
for decreasing the occurrence of a problem behavior must be implemented
instead (see Chapters 15–18).

Can Consistency Be Maintained? For extinction to be implemented correctly,
the reinforcer must never follow the problem behavior. This means that all people
involved in the treatment must be consistent and eliminate the reinforcing conse-
quence each time the problem behavior occurs. If the problem behavior is reinforced
even occasionally, the procedure amounts to intermittent reinforcement for the
behavior rather than extinction. Lack of consistency is a common reason for the fail-
ure of extinction procedures (Vollmer, Roane, Ringdahl, & Marcus, 1999). For
example, if parents are implementing extinction consistently for their child’s bedtime
tantrums, but the grandparents occasionally reinforce the problem when they visit,
the tantrums will not be eliminated. Likewise, if most staff members implement
extinction for Willy’s refusals and sexist comments, but one or more of them
continue to pay attention to the behavior, the behavior will not be eliminated.

All change agents must be trained to use the procedure correctly to ensure
consistency in implementing an extinction procedure. The change agents must
receive clear instructions to be consistent and a rationale explaining why consis-
tency is important. Furthermore, the best results are achieved if the extinction
procedure is modeled for the change agents and they have the opportunity to
rehearse the procedure and receive feedback. In some cases, it is beneficial to
have contingencies of reinforcement for the change agents’ correct use of the
extinctio