Association for Behavior Analysis International

One of the enterprises of a science of behavior is to generate an end product (data) that faithfully represents a quantitative description of natural or imposed variability characteristic of events. Our success at this over the years is undeniable and has served as the basis for describing the emerging practices from the science as "evidence-based". The manner in which we "announce" our evidence falls under the rubric of data display, the process of transforming data from investigation or treatment effects into stimuli whose function is to influence interpretive behavior. The behavior analyst is in a unique position to assist other professionals in interpreting their own body of facts so that they can reap the benefits of their respective science and practice. During early developmental stages, applied behavior analysis utilized cumulative curves to describe important relationships between independent and dependent variables. Examples from these early reports serve as powerful demonstrations of an emergent technology of behavior (Wolf, Risley, & Mees, 1964). Oddly enough, the practice of presenting cumulative curves to describe data has virtually disappeared in applied behavior analysis, and our departure from this practice seems to have occurred without notice or discussion—a premature and perhaps a regrettable outcome of steady state research. In this symposium, three papers are presented in which data are reported using cumulative curves. A close examination of these curves reveals behavior-environment interactions that may have been neglected using traditional graphing procedures. The implications of using cumulative curves will be discussed, as well as further directions in function-based treatments, and in applied behavior analysis, in general.

Professionals have a vested interest in the display of behavior graphs in as much as they can demonstrate the relative contributions from various interventions. Behavior analysts developing treatment plans often work closely with psychiatrists who are addressing behaviors pharmacologically. Medication adjustments involve manipulations of independent variables and these effects are of paramount importance to the outcome of treatment. This presentation addresses the use of cumulative graphing techniques to determine the effects of medication adjustments on aggressive behaviors that are concurrently treated by withdrawing attention contingent upon aggressive behaviors (e.g.,: attention), and providing contingent delivery of conditioned reinforcement for alternative behaviors. Three subjects between the ages of 35 and 50 were admitted for treatment of aggressive behaviors after suffering a brain injury. Functional behavior assessments suggested that aggressive behaviors by subjects had been shaped and maintained by attention delivered following these behaviors. Differential reinforcement procedures were developed and implemented approximately 24–72 hours after admission, and remained in effect throughout the reported treatment period. Medication adjustments also served as independent variables. Figure 1 shows a graph using a more traditional plotting method in which aggressive behaviors were recorded daily. Figure 2 presents these same data plotted in a cumulative graph. The2 curves are generally negatively accelerated, and contain within them "mini" extinction curves, which developed primarily when medications were decreased. Two other subjects received the same general treatment and medication adjustments, and cumulative curves generated were essentially the same in shape. Implications will be discussed, both from the perspective of a behavior analyst and a psychiatrist.

Social networking websites are becoming more popular and more easily accessible for all populations. The online social networking provides difficulty for a controlled environment to directly address inappropriate attention-seeking behavior. The research in this study is relevant to this population as individuals are being provided with attention for making inappropriate posts; such as, threats of self-harm, solicitation of sex, and inappropriate false reporting of caregivers. The subject of this study was a 26-year-old individual with a diagnosis of acquired brain injury. A multielement treatment reversal design with cumulative graphing was utilized during the study. The present study introduced3 confederates who were already established as caregivers and members of the subject's social networking website (Facebook). The confederates were provided simple instructions of how to implement differential reinforcement of alternative behavior within the confines of Facebook. In the presented research, evidence supports a treatment effect for reducing the number of inappropriate attention seeking behavior and increasing the appropriate attention seeking behavior through the establishment of confederates trained in providing differential reinforcement of alternative behavior in a social networking website.

Behavior treatment plans often use differential reinforcement strategies that withhold reinforcement for target behaviors, and provide reinforcement for alternative behaviors. The relation between the specific target behaviors and designated alternative behaviors are usually derived from functional assessment or functional analysis procedures. The alternative behaviors identified in these relations have been referred to as "replacement behaviors," but this term has not been adopted uniformly byapplied behavior analysispractitioners and there are compelling reasons to discard this term. In this presentation,2 case studies are presented in which behaviors targeted for reduction were determined to be related to attention from caregivers. Cumulative graphs were used to analyze differential reinforcement interventions. Following apparent early treatment success, negatively accelerating trends for target behaviors reversed abruptly following inadvertent positive reinforcement (attention) for target behaviors; however, positively accelerating curves for alternative behaviors remained unchanged. Figures 1 and 2 show cumulative responses for both subjects. Such a reversal is not uncommon, but rather reveals that differential reinforcement contingencies can be transient and brittle, often reflecting ongoing fluctuations in competing reinforcement contingencies. Under these circumstances, the term "replacement behaviors" is misleading and could generate misconceptions about the enduring nature of ABA treatments. Other terms are suggested and the use of cumulative recording is encouraged to track changes in competing reinforcement contingencies.