A Comparison of Single-Case Effect Measures Using Check-In Check-Out Data.

Abstract

There are numerous effect measures researchers can select when conducting a meta-analysis of single-case experimental design research. These effect measures model different characteristics of the data, so it is possible that a researcher's choice of an effect measure could lead to different conclusions about the same intervention. The current study investigated the impact of effect measure selection on conclusions about the effectiveness of check-in check-out (CICO), a commonly used intervention within School-Wide Positive Behavior Interventions and Supports. Using a multilevel meta-analysis of seven different effect measures across 95 cases in 22 studies, findings suggested that all effect measures indicated statistically significant results of CICO in improving student behavior. However, the magnitude of the effects varied when comparing the results to interpretive guidelines, suggesting that the selection of effect measures may impact conclusions regarding the extent to which an intervention is effective. Implications, limitations, and future directions are discussed.