Objective: This paper demonstrates how to conduct a meta-analysis that includes both between-group and single-case design (SCD) studies. The example studies whether choice-making interventions decrease challenging behaviors performed by people with disabilities.
Methods: We used a between-case d-statistic to conduct a meta-analysis of 15 between-group and SCD studies of 70 people with a disability, who received a choice intervention or control. We used robust variance estimation to adjust for dependencies caused by multiple effect sizes per study, and conducted moderator, sensitivity, influence, and publication bias analyses.
Results: The random-effects average was d = 1.02 (standard error of 0.168), so the 95% confidence interval (CI) suggests choice-making reduces challenging behaviors by 0.65 to 1.38 standard deviations. Studies that provided choice training produced a significantly larger intervention effect.
Conclusion: Choice-making reduces challenging behaviors performed by people with disabilities.
Keywords: Meta-analysis; challenging behaviors; choice; disabilities; robust variance estimation.