Identifying genuine effects in observational research by means of meta-regressions

Meta-regression models are increasingly utilized to integrate empirical results across studies while controlling for the potential threats of data-mining and publication bias. We propose extended meta-regression models and evaluate their performance in identifying genuine empirical effects by means of a comprehensive simulation study for various scenarios that are prevalent in empirical economics. We can show that the meta-regression models here proposed systematically outperform the prior gold standard of meta-regression analysis of regression coefficients. Most meta-regression models are robust to the presence of publication bias, but data-mining bias leads to seriously inflated type I errors and has to be addressed explicitly.