David B. Wilson, PhD, George Mason University
Version date: 2023.11.27
Binary Dependent Variable: Proportions within Each Condition with Pretest Data
Notes: This method approximates Cohen’s d and Hedges’ g from binary or dichotomous outcome data using the proportion of participants in each condition experiencing the outcome at pretest (baseline) and post-test (follow-up). The method produces a difference-in-differences effect size estimate. Two computational methods are provided, the logit method and the Cox logit method. If a proportion equals 0, the proportion of successes is set to .5/n, where n is the sample size for that condition. Enter the correlation between the pretest/post-test values if known. If unknown, a default value of 0.50 will be used. This value does not affect the effect size but does affect the standard error/variance. As such, the standard error is an approximation.

This work is licensed under a Creative Commons Attribution-Noncommercial 3.0 United States License.
