Evaluation Methods for Intervention Effect in Single-Case Experimental Design

Abstract

Single-case experimental design (SCED) is a research design used to evaluate the efficacy of interventions on individuals and groups. In this paper, we present an overview of visual analysis, the primary method for testing the efficacy of SCED, and discuss issues related to its reliability. To assist visual analysis, we explained the procedures of structured visual analysis method using visual aids and statistical measures. For within-case statistical measures, we explored Tau-based measures based on overlapping of data, mean differences between phases and log response ratio (LRR), and regression model-based methods. Regarding between-case effect measures, we investigated a meta-analytic integration method of standard mean differences between cases and between-case standard mean difference (BC-SMD), based on hierarchical linear models. Finally, we discussed the guidelines for selecting appropriate statistical indicators and introduced user-friendly software and web applications for conducting and calculating visual/statistical indicators.