Using linear mixed effects models: A single-case experimental design meta-analysis of functional communication training

Abstract

In this study, we extended the meta-analytic literature on single-case experimental design (SCED) studies on functional communication training (FCT) using a linear mixed effects model (LMEM). First, we evaluated the methodological adequacy of each study using the What Works Clearinghouse (WWC) Standards. Second, we calculated a standardized rate difference (SRD) between the rates of alternative communicative responses and problem behavior. Third, we estimated a LMEM using SRD as the dependent variable to examine the effect of FCT across participants with different disabilities. Finally, we estimated a LMEM including only studies in which participants had a diagnosis of Autism Spectrum Disorder (ASD), without standardizing the rate difference. Results indicate that FCT was effective in increasing the SRD as suggested by the greater number of alternative communicative responses compared to the number of problem behaviors over time. Results also suggest that FCT is an effective intervention for participants with ASD and that it might not be necessary to standardize scores when using LMEMs to analyze data from participants with the same disability. We discuss the utility of analyzing SCED data using LMEMs and make recommendations for an SCED data warehouse protocol.

Keywords:

• Meta-analysis
• Single-case experimental design
• Linear mixed effects models
• Functional communication training