Meta-Regression-Based Meta-Analysis

Meta-regression-based meta-analysis extends standard meta-analysis by fitting a weighted regression model in which study-level characteristics (moderators) predict observed effect sizes. Rather than simply pooling effects, this approach asks why effects vary across studies — linking heterogeneity in outcomes to differences in population, intervention, design, or measurement features. It is the primary tool for explaining between-study variance in quantitative evidence synthesis.