Applying Location Scale Models to Meta-Regression in Education

• In Event: Advancing Meta-Analytic Methods

Abstract

The present study introduces location scale models to meta-regression in education research, demonstrating how moderators can affect both the average effect sizes (location) and their variability (scale). Using two empirical meta-analyses in education, we show how factors such as socioeconomic status influence not only mean outcomes but also heterogeneity across studies. In addition, we discuss limitations of the current modeling approaches, including extensions to multilevel and multivariate frameworks. By applying this advanced framework, we aim to promote more nuanced interpretations of heterogeneity and encourage broader adoption of distributional methods in educational meta-analysis.

Authors

• Bixi Zhang, Graduate Center - CUNY

• Mali Waugh, Graduate Center - CUNY

• Eduardo Alarcon, Graduate Center - CUNY