A Comparison Among Different Approaches to Data from Single Subject Designs

摘要

Both practical and theoretical concerns motivate carefiil evaluation of treatment programs for aphasia and related disorders. A common paradigm for interventions at preliminary stages of development is based on data from individual patients. These designs typically include a baseline phase during which several observations are made of some stable target behavior, followed by an intervention. The question is whether an important diftrence in behavior can be attributed to the intervention. Single subject designs can provide methodological rigor and flexibility. However, there is uncertainty concerning the most appropriate approaches to data analysis. This paper outlines conceptual and statistical challenges and, then, compares two novel approaches to data analysis, (1) a combination of multiple regression and time series procedures and (2) a bootstrapping/simulation approach.