|
| Recent Advances in Treatment Integrity Research |
| Sunday, May 24, 2009 |
| 9:00 AM–10:20 AM |
| North 121 A |
| Area: EDC; Domain: Applied Behavior Analysis |
| Chair: Florence D. DiGennaro Reed (Melmark New England) |
| CE Instructor: Gerald E Harris, Ph.D. |
| Abstract: Treatment integrity refers to the extent to which an intervention is implemented as designed and is an area of research that has increasingly become more common. This symposium highlights four talks that share recent advances in treatment integrity research. The first talk presents findings from a retrospective analysis of archived data to examine temporal patterns of data collection. In the second presentation, the effect of individualized video training alone and in combination with performance feedback on the integrity of behavior plan implementation will be shared. Next, the impact of types of integrity failures (e.g., errors of commission and omission) on academic behavior of students will be presented. The symposium concludes with a talk that presents results of an investigation of a variation to the DRA procedure that might be useful when high treatment integrity is not possible given clinical concerns. Presenters will discuss implications of their findings in educational and clinical settings. |
| |
| Behavioral Assessment of Treatment Integrity Data Collection: Considerations Regarding Temporal Sequencing |
| DEREK D. REED (Melmark New England), Daniel Mark Fienup (The May Institute), Gary M. Pace (The May Institute) |
| Abstract: With rising interest in the role of treatment integrity on student outcomes, research has primarily focused on isolating the techniques and procedures necessary to improve staff’s acquisition and maintenance of adequate levels of integrity. Despite increasing numbers of publications on this topic, there is little discussion of the variables surrounding the collection of integrity data. Using an archived database of logged integrity checks at a residential school for children with brain injury, we sought to examine the degree to which integrity data collection conformed to the best practices of general behavioral assessment (due to the use of an archival data source, reliability measures were not possible) with respect to temporal sequencing. Moreover, due to the agency’s goal of collecting integrity on each student per month, we sought to whether the sequencing of integrity checks scalloped in conformity to fixed-interval responding. Results indicated that 50% of the staff exhibited scalloping in their collection of integrity data. We discuss the possible stimulus control and reactivity on the part of the teachers being observed when integrity checks are conducted in scalloped patterns. We conclude with a discussion on possible procedures to support the distributed collection of integrity data in applied setting. |
| |
| Effects of Video-training on Treatment Integrity of Behavior Support Plans |
| FLORENCE D. DIGENNARO REED (Melmark New England), Robin Codding (University of Massachusetts Boston), Cynthia N. Catania (Melmark New England), Helena L. Maguire (Melmark New England) |
| Abstract: A number of studies have examined ways to promote accurate behavior support plan implementation and have generally shown that teachers require ongoing support and performance feedback to maintain high levels of accuracy. These studies have focused on follow-up strategies to address poor treatment implementation and questions regarding how best to train teachers initially remain unanswered. The purpose of the present study is to examine the effects of individualized video-training (IVT) on the accurate implementation of behavior support plans in a setting that treats individuals with significant behavior problems. Using a multiple baseline design across 3 teachers, findings revealed that IVT improved treatment integrity above baseline levels; however, teacher performance remained variable. The addition of verbal performance feedback (PF) increased treatment integrity to 100% across 3 consecutive sessions for all participants. Performance was maintained at a one-week follow-up probe. Social validity data suggested that teachers found the IVT+PF the most acceptable procedure. |
| |
| An Evaluation of Integrity Failures during Token Economies |
| ALLISON TETREAULT (West Virginia University), Claire St. Peter Pipkin (West Virginia University), Timothy R. Vollmer (University of Florida), Shari Marie Winters (West Virginia University) |
| Abstract: Token economies are empirically supported behavior management systems, but may be implemented imperfectly by teachers. Although research has examined the effects of incorrect implementation (treatment integrity failures) on differential reinforcement in general, little is known about the effects of treatment integrity failures during token economies. In the current study, we examined the impact of errors of commission (unscheduled reinforcers are delivered) and omission (scheduled reinforcers are not delivered) on academic behavior maintained by token economies with 6 typically developing students. Students were exposed to baseline, a token system with full integrity (implemented perfectly) and reduced levels of integrity with both errors of omission and commission in a multiple-baseline-across-participants design. During reduced integrity phases, the system was implemented correctly 50% of opportunities, because this level seemed to be possibly detrimental to differential reinforcement interventions in previous research. Effects of token systems were not detrimentally affected by implementation at 50% treatment integrity (with either omission or commission). These findings suggest that token systems may be recommended when levels of treatment integrity are a concern. |
| |
| Parametric Evaluation of the Differential Reinforcement of Alternative Behavior Procedure |
| ELIZABETH S. ATHENS (ABA Learning Centre), Timothy R. Vollmer (University of Florida) |
| Abstract: Differential reinforcement of alternative behavior (DRA) is a commonly used treatment for problem behavior. Usually with DRA, problem behavior is placed on extinction while an alternative, more appropriate behavior, is reinforced. In some cases, however, the targeted problem behavior is too disruptive or dangerous to place on extinction. In such cases, DRA cannot be implemented with consistent treatment integrity. The purpose of the current study, therefore, is to evaluate a variation of the DRA procedure that did not include an EXT component and was designed to provide more immediate, longer duration, and higher quality reinforcers for appropriate behavior relative to reinforcers for problem behavior. To do this, we differentially manipulated the parameters of reinforcement along several dimensions. Specifically, for the appropriate response (relative to the inappropriate response) we made a) reinforcement following this behavior more immediate b) the duration of reinforcement greater c) increased the quality of reinforcement. The effects of these isolated manipulations became clinically valid when a combination of each favored appropriate behavior in the final manipulation. Under such final manipulations, differential reinforcement effects were obtained with individuals who engaged in socially reinforced severe problem behavior. The procedure is conceptualized as differential reinforcement insofar as reinforcement parameters differentially favored appropriate behavior. |
|
| |