Effect Size Benchmarks for Response Elaboration Training: A Meta-Analysis

Purpose With a number of single-case experimental design studies reporting the effects of treatment for response (and modified response) elaboration training (RET/M-RET), it is important to consolidate data over multiple participants to allow comparison within/between individuals and across similar treatments. The purpose of this study was to conduct a meta-analysis of single-case experimental design studies of RET/M-RET and to determine effect size (ES) benchmarks to allow comparison to "group" data. Method Database and bibliographical searches identified 20 investigations of RET/M-RET. Nine studies had sufficient experimental quality, compliance with the essential components of the RET protocol, and consistency in the dependent variable (i.e., accurate content production in response to picture stimuli) to be retained for the meta-analysis. Probe data for a total of 26 persons with aphasia (PWA) were extracted from published graphs (if raw data were not available) to calculate weighted ESs at the end of treatment and at follow-up for both treated and untreated stimuli. The first, second, and third quartiles of the distributions were used to serve at benchmarks for small, medium, and large effects. Results Nearly all participants demonstrated positive effects as a result of RET/M-RET, indicating an association with positive changes in content production for PWA. Small, medium, and large benchmarks are reported for treated items after treatment and at follow-up, as well as for untreated items after treatment and at follow-up. Conclusions With a larger sample of 26 participants, this analysis indicates that RET/M-RET are associated with positive changes in content production for PWA. ES benchmarks allow clinicians/researchers to compare an individual's performance across multiple applications of treatment to performance of other PWA and to other treatments with similar outcomes.