Design and Analysis of Single-Case Research

T. Mark Beasley School of Education and Human Services St. John's University, New York

David B. Allison Obesity Research Center St. Luke's/Roosevelt Hospital Columbia University, College of Physicians & Surgeons

Bernard S. Gorman Department of Psychology Hofstra University, Nassau Community College

Among behavior analysts, single-case designs have become a popular alternative to more traditional group experiments. Although single-case designs have numerous strengths both philosophically and practically, they also have potential weaknesses in the areas of analysis and inference. In attempts to address a major analytic issue (i.e., the limited numbers of observations), the use of frequent and repeated measures is often suggested ( Barlow & Hersen, 1984). The collection of data at several points in time, however, creates the potential for "testing" and "instrumentation" ( Campbell & Stanley, 1966) as threats to internal validity ( Horn & Heerboth, 1982). The primary focus of this chapter is that undetected cyclical trends in behavior (or factors correlated with the target behavior) may confound estimates of treatment effects. Moreover, cyclicity, even when detected, complicates the interpretation of data and remains one of the most neglected issues in single-case research ( Barlow & Hersen, 1984).

Cyclicity can be defined as any recurrent patterns or fluctuations in

-279-