Standardized Mean Difference Effect Sizes in Medical Education: A Review and Recommendations (Poster 49)

• In Event: AERA Poster Session Nine
  In Poster Session: Division I Poster Session

Abstract

The What Works Clearinghouse provides guidance when using standardized mean difference effect sizes, including: (1) Hedges’s g should be reported in favor of Cohen’s d, particularly when sample sizes are small (<50), (2) Cohen’s conventions for small, medium, and large effects of 0.2, 0.5, and 0.8, respectively, are strongly recommended against, even by Cohen himself, and (3) the improvement index is a useful means of describing intervention effects and is recommended for use. Journals publishing medical education research are reviewed for effect size reporting practices. The review indicates that Hedges’s g is not used, Cohen’s conventions are frequently relied upon, and the improvement index is not used. Recommendations are made to improve these practices.

Author

• Peter Boedeker, Baylor College of Medicine
  Presenting Author