Examining Computer-Based Scaffolding Research Quality Through a Risk of Bias Lens

• In Event: 52.078 - Recent Reviews in Education

Abstract

Baeysian Network Meta-Analysis represents a rather unique challenge in assessing the quality of included studies. Prior efforts to synthesize computer based scaffolding are in need of a closer examination of research quality. This study examines two quality metrics for meta-analysis, study design, and risk of bias (Higgins et al., 2011). Lower quality study designs are associated with greater learning gains. A targeted examination of specific risks suggests the opposite. Low risk of bias is associated with higher gains (random sequence g = 0.23; selective reporting g = 1.13) than studies with a high risk. The only consistent predictor of learning for study design or risk of bias is being part of a computer based scaffolding treatment.

Authors

• Andrew Walker, Utah State University

• Brian R. Belland, Utah State University

• Nam Ju Kim, University of Miami

• Mason Lefler, Utah State University