Accounting for sample overlap in economics meta‐analyses: The generalized‐weights method in practice

Abstract

Meta‐analyses in economics frequently exhibit considerable overlap among primary samples. If not addressed, sample overlap leads to efficiency losses and inflated rates of false positives at the meta‐analytical level. In previous work, we proposed a generalized‐weights (GW) approach to handle sample overlap. This approach effectively approximates the correlation structure between primary estimates using information on sample sizes and overlap degrees in the primary studies. This paper demonstrates the application of the GW method to economics meta‐analyses, addressing practical challenges that are likely to be encountered. We account for variations in data aggregation levels, estimation methods, and effect size metrics, among other issues. We derive explicit covariance formulas for different scenarios, evaluate the accuracy of the approximations, and employ Monte Carlo simulations to demonstrate how the method enhances efficiency and restores the false positive rate to its nominal level.