Paper 217-25 Robust Effect Size Estimates and Meta-Analytic Tests of Homogeneity

This paper presents a computer program that calculates two effect size estimates (gamma and trimmed-d) that are robust to violations of the assumptions of population normality and homogeneity of variance. Because of their robustness properties, these indices are frequently more appropriate than the conventional parametric indices such as Cohen’s d or Hedges’ g. Additionally, a program that uses these effect size indices in robust permutation tests for betweenclass homogeneity (a test frequently used in metaanalysis) is introduced. The paper provides a demonstration of the SAS/IML‚ code and examples of the application of the code in simulation studies.