Association for Behavior Analysis International

Alternating treatments designs can be used to evaluate multiple interventions and compare interventions. This presentation will address common interpretation errors that are associated with standard alternating treatments designs and propose solutions for each type of error. First, the presentation will focus on how researchers frequently conclude that an intervention or multiple interventions were effective, when changes may have been caused by uncontrolled threats to internal validity. A design solution to this problem, including a no-treatments series during the alternating treatments phase, will be described and analyzed. Next, the presentation will focus on misinterpretation associated with cumulative learning data. A proposed solution to this problem focus on supplementing repeated measures cumulative learning figures with figures that plot learning per session data. Discussion focuses on applied strengths of alternating treatments designs, effect size analysis, and how interpretation errors can adversely affect consumers of applied science.

researchers, graduate students, consumers of research

Christopher H. Skinner received his Ph.D. in School Psychology program from Lehigh University in 1989. While at Lehigh, he served as a special education teaching assistant for elementary students with Autism and as a teacher for 10th grade students with emotional/behavioral disorders. After finishing his Ph.D., Skinner was an assistant professor at The University of Alabama (3 years) and coordinator of School Psychology Programs at Mississippi State University (7 years) and The University of Tennessee (15 years). Currently, he is Professor at The University of Tennessee and teaches graduate courses in the School Psychology and Behavior Analysis programs. Skinner has co-authored over 200 peer-refereed journal articles and earned three national research awards including the Fred S. Keller Award for empirically validating interventions. Skinner’s accomplishments can be directly traced to his work with exceptional graduate students and practitioners. He is happiest when he is partnering with his students to work with educational professional to remedy presenting behavior or academic problems, while also conducting applied research. To simultaneously accomplish these goals, Skinner and his collaborators have relied single-subject design procedures.