Research Methods in Applied Behavior Analysis

Jon S. Bailey, Mary R. Burch

• Publisher: Routledge
• Year: 2017
• ISBN:
  9781134859207
• Edition: 2
• Topic: Psicología
• Subtopic: Formación en psicología

Research Methods in Applied Behavior Analysis, 2nd edition, is a practical text that provides the beginning researcher with a clear description of how behavior analysts conduct applied research and submit it for publication. In a sequence of ten logical steps, the text covers the elements of single-case research design and the practices involved in organizing, implementing, and evaluating research studies. This revision includes new material on how to critique a journal article and how to extract treatment ideas from published studies. The novice researcher receives a "steps for success" approach that is brief, useful, to the point, and clearly delineated.

Part I

A Brief Introduction to Behavior Analysis Research

Applied research is a creative and demanding activity that requires critical thinking, insight into multiple issues associated with behavior, and an extraordinary ability to be doggedly consistent. The successful applied researcher must also have the social skills required to work carefully with others and be able to motivate everyone on the research team. Having these critical skills can make or break a research project. When everything comes together—the technical skills, social skills, consistency and determination—the researcher can play an important part in making socially significant changes for the community, and powerful, clinically significant changes in the lives of individuals.

The field of applied behavior analysis has attracted a great deal of attention in recent years as the promise of an effective technology based on the products of an experimental analysis of behavior (Morgan & Morgan, 2001; Skinner, 1974) has begun to emerge. The contributions of the field are well documented in numerous journals now specifically devoted to the topic and in numerous excellent texts and collections of readings. This extensive literature describes the important contributions that have been made in autism spectrum disorders, mental retardation, rehabilitation, delinquency, mental health, counseling, education, business and industry, and many other fields. It represents the remarkable development of practical procedures for behavior change arising from the experimental analysis of behavior with lower organisms.

Socially significant target behaviors are identified, functions are identified, interventions are revised and put in place, and the outcome is important to consumers and their families alike.

Applied behavior analysis has attracted a multitude of enthusiastic participants, in part due to an apparent but deceptive simplicity: Socially significant target behaviors are identified, functions are identified, interventions are revised and put in place, and the outcome is important to consumers and their families alike. The apparent simplicity of the treatments may actually be a function of a limitation on journal space that prevents researchers and authors from fully describing the extensive efforts necessary to accomplish the dramatic behavior. Published research articles almost never describe how a researcher came up with the original idea, nor do they trace the evolution in thinking that occurred through stimulating brainstorming sessions, sharing data with colleagues over coffee, or being lucky enough to be in the right place at the right time. Published articles rarely detail how the researchers happened to locate the environment where they work, gained the confidence of the established authority, and obtained permission to carry out their study. The specific details involved in the moment-to-moment conduct of the research once it is under way are simply never discussed in the text of an article. Such statements as, “five preschool children were chosen” or “the study was carried out in three nursing homes” simply do not properly describe the extensive planning, intensive strategizing, as well as social and diplomatic behavior required by the applied researcher. The abridged public version of studies, driven by the cost of journal space, grossly misrepresents the sustained effort and finely tuned skills researchers must acquire to conduct quality experimentation in applied settings.

The abridged public version of studies, driven by the cost of journal space, grossly misrepresents the sustained effort and finely tuned skills researchers must acquire to conduct quality experimentation in applied settings.

Novice researchers, without the advantage of having apprenticed with established faculty mentors, must often engage in extensive, frustrating, and wasteful trial-and-error activity trying to figure out how our studies are carried out so easily. The inside story is that applied behavioral research can be exceedingly difficult to accomplish, often much more so than basic laboratory research where the subjects come in packing crates from a breeding farm, the experimental equipment is off the shelf, the established protocols are already memorized and the research questions are derivative. The careful preplanning, eye for detail, compulsion for precision and consistency and abundant persistence required of all successful researchers are obviously necessary. However, the creative, managerial and administrative skills needed to conceptualize a unique applied problem, scale it in proportion to available resources, and orchestrate the applied research are not obvious to the novice researcher.

The present text is an effort to fill this void in the applied behavior analyst’s repertoire. Over the years, it has become clear that there is a proven formula for successful research. A major goal here is to describe that formula for aspiring researchers and practitioners in the form of helpful hints, suggested guidelines and detailed specific steps for conducting a research project. Many of the recommendations are the result of the first author’s nearly 50 years of experience of guiding and directing theses and dissertations at Florida State University. The intent of this text is not to review the literature or prescribe treatment procedures, but rather to give eager new researchers a set of guidelines by which they may carry out their own initial research and hopefully contribute to the science of behavior.

Learning to “separate the wheat from the chaff” is a vital skill for the ethically-minded practitioner.

In addition, since many newly minted behavior analysts will only be consumers of this research, we are adding two additional chapters. One of these chapters involves critiquing published studies to determine if they meet the standards of the Behavior Analyst Certification Board: “Behavior analysts always have the obligation to advocate for and educate the client about scientifically supported, most-effective treatment procedures.” For many, this simply means base your treatments on published research, but the standards for publication vary greatly from one journal to the next. Learning to “separate the wheat from the chaff” is a vital skill for the ethically minded practitioner. Second, we are adding a chapter on how to derive the essential, reliable, effective procedures from the resulting pre-screened research. Before we proceed, however, a brief review of some basic features of operant research and a few landmark studies that have contributed to our current methodology seems to be in order.

The Operant Research Model

The earliest applied research studies were patterned in many ways after the style of research developed by B. F. Skinner and his students (Ferster & Skinner, 1957; Skinner, 1938) as they worked with rats and pigeons. A fundamental feature of the methodology is the detailed analysis of the behavior of the individual organism. As was the case with the original laboratory studies, a small number of individual participants (very often only one) were observed for extensive periods of time. In these early, applied studies, the “free-operant,” ongoing behavior (frequently dangerous or deviant in nature) of one or at most a handful of human participants was measured during an initial baseline condition to determine the operant level of performance. Subsequently, some type of intervention procedure was systematically brought into contact with the target behavior, and the procedure’s effects were noted. At some later point in these original studies, the contingent procedures were withdrawn to provide further evidence of the procedure’s functional relationship to the target behavior.

Although this seemingly simple extrapolation of a kind of “case study” methodology would seem to be an easy extension, the task required some adjustments in experimental methods. New data-recording techniques had to be developed in the transition from laboratory to field research environment. Experimental participants were humans with broader behavioral repertoires and larger environments. Manipulated consequences were more than food pellets and two-second presentations of food hoppers. Target behaviors having social importance were usually not amenable to measurement with automated devices. Instead, human observers had to be extensively trained, and ways of ensuring the objectivity of measurement procedures were necessary. The reintroduction of the human observer into experimental analysis brought with it a host of logistical and methodological problems. These concerns were added to the difficulties associated with attempts to stabilize experimental environments in field settings. All of this contributed to the difficult early development of the emerging field of applied behavior analysis.

Important Methodological Milestones

Although it is often taken for granted, systematic observation procedures for collecting data on significant human behaviors in natural settings are a relatively recent development, which we will explain in some detail.

Measurement of Significant Behavior in Applied Settings

The seminal article “The Psychiatric Nurse as a Behavioral Engineer” by Ayllon and Michael (1959) represented a major contribution to applied behavior analysis. In this earliest of studies, problem behaviors of chronic psychotic patients were selected, defined, and measured systematically. In addition to the straightforward counting of such behaviors as “number of entrances to the nurses’ office” or “number of magazines hoarded,” Ayllon and Michael employed time-sampling recording procedures to evaluate “psychotic talk” or the presence of nonviolent behavior at 15- to 30-second intervals during the experimental sessions. The application of this efficient and accurate method of behavior measurement represented a major innovation in applied human research. Various aspects of this method of behavior measurement were, of course, part of human factors engineering in industrial psychology, but the more precise application of regular observation techniques to applied human research represented a largely new development. The reversal design, which is described in detail in a later section of this text was also successfully applied by Ayllon and Michael to the hospital ward environment. Figure 1.1 is taken from this published report. This graph reveals that an initial pretreatment condition consisting of one session was subsequently followed by a reinforcement condition and then later by a return to the pretreatment situation—the original ABA design.

Refinement of Observation and Control Procedures

Over the next few years, applied researchers began applying this new technology more extensively to modify deviant child behavior. During the 1960s, researchers such as Don Baer, Mont Wolf, and Sidney Bijou, who were all at the University of Washington at the time, made significant advances relative to changing the behavior of “deviant” children. One of the first refinements came in the area of observation procedures, where a more intensive analysis of moment-to-moment changes in behavior was needed. The 10-second-interval recording method evolved and was an important advance in the precision of behavior observations (see Allen, Hart, Buell, Harris, & Wolf, 1964). Figure 1.2 shows one of the first widely distributed descriptions of this kind of observation technique.

Figure 1.1 Reinforcement and subsequent extinction of the response “being on the floor.”

Source: “The Psychiatric Nurse as a Behavioral Engineer,” by T. Ayllon and J. Michael, 1959, Journal of the Experimental Analysis of Behavior, 2, pp. 323–334. Copyright 1959 by the Society for the Experimental Analysis of Behavior. Reprinted with permission.

Figure 1.2 The original 10-second-interval observation form and code.

Source: From “The Effects of Social Reinforcement on Isolate Behavior of a Nursery School Child,” by E. Allen, B. Hart, J. Buell, F. Harris, and M. Wolf, 1964, Child Development, 35, pp. 511–518. Copyright 1964 by the Child Development Journal. Reprinted with permission.

Allen et al. (1964) employed other important advances in data analysis and methodology. Their study elaborated the use of the reversal design by extending the original baseline period over several sessions. This allowed for a better analysis of the inherent variability of pre-intervention behavior and made possible an analysis of possible trends in the performance from day to day. In addition, the added refinement of post-treatment checks gave evidence of the durability of behavior change (Figure 1.3). The Allen et al. study also presented evidence of inter-observer agreement (IOA) to demonstrate that the definitions of behavior were clear and that the features of behavior that were selected by the investigators were, in fact, the major controlling stimuli for the observer. Although today IOA procedures are taken for granted, the implementation of their use in studies such as the one by Allen et al. (1964) typified the movement toward greater precision in behavior analysis in its early years.

The mid-1960s represented a period of exciting advances in the infant technology of applied behavior analysis. The Allen et al. study was accompanied by several other important publications that used similar methodological refinements (see Harris, Johnston, Kelley, & Wolf, 1964; Hart, Allen, Buell, Harris, & Wolf, 1964). These and other studies were the outgrowth of the laboratory-based child research programs described by Bijou and Baer (Bijou, 1961; Bijou & Baer, 1961; Bijou & Orlando, 1961; Bijou & Sturges, 1959) and verified the approach to child development taken by these researchers.

Figure 1.3 Percentages of time spent in social interaction during approximately 2 hours of each morning session.

Source: “The Effects of Social Reinforcement on Isolate Behavior of a Nursery School Child,” by E. Allen, B. Hart, J. Buell, F. Harris, and M. Wolf, 1964, Child Development, 35, pp. 511–518. Copyright 1964 by the Child Development Journal. Reprinted with permission.

Science and Clinical Success

The well-documented and substantial changes in behavior reported in the early published studies suggested that operant methodology would ultimately play a major role in the development of a behavioral technology for treating psychopathology.

Perhaps the most dramatic and well publicized of this early series of applied studies involved a 3½-year-old autistic child named Dicky. Dicky had surgery to remove the cataract-clouded lenses of his eyes, and he was in danger of losing his vision if he did not begin wearing prescription glasses (Wolf, Risley, & Mees, 1964). This research program clearly revealed that scientific rigor need not be sacrificed during design of the successful treatment of severely deviant behavior. The program involved the strategy of breaking a complex repertoire into smaller components (tantrums, self-destructive responses, bedtime problems, wearing glasses, throwing glasses, eating problems) and gaining control over each one separately. The result was that the previous...