ABSTRACT
The chapter explains how to conduct high-validity case studies and single-case experiments that use objective measurement techniques, multiple dependent variables, multiple sources of information for each variable, frequent assessment of the variables, multiple replications using heterogeneous participants, and control cases that test both convergent and discriminant hypotheses. The importance of minimizing history and maturation confounds is reviewed as is the need to choose a high-impact independent variable that can be standardized and that has a long-term effect. If possible, an independent variable should be manipulated. It is important to carefully consider the case that is selected and the appropriate unit of analysis for that case (such as group as a whole versus individual group members). The chapter explains good data collection procedures, including having clear operational definitions, well-defined sources, and a written chain of evidence. The chapter reviews the five basic single-case experimental designs: A-B-A, A-B-C-B, multiple baseline, simultaneous treatment, and changing criterion. For all designs, it is important to establish a stable baseline. Both qualitative and quantitative single-case designs are discussed, as are methods for interpreting results from single-case designs.
