© 2019 Modern Meta-Analysis Research Institute. Proudly created with Wix.com

ABOUT

Modern meta-analytic methods are needed to understand the conditions necessary to maximize STEM intervention impacts on learning outcomes. Meta-analysis is a suite of techniques that is uniquely positioned to answer important questions about contextual factors related to intervention effects. Unfortunately, traditional applications of meta-analysis in STEM, and education research broadly, focus on identifying average effects across multiple studies. Such applications are helpful but they fail to provide policy-makers and practitioners concrete guidance when effects vary substantially across those studies (i.e., effect heterogeneity). To answer questions related to effect heterogeneity, STEM researchers conducting meta-analyses need a broader set of modern analytical tools, such as methods for handling correlated effect sizes, effect sizes from complex sampling designs, and adjusting for complex patterns of reporting bias.

‚Äč

The purpose of this 5-day workshop is to provide STEM researchers with a comprehensive, introductory meta-analysis workshop focused on state-of-the-art methods including use of the program R. Like other introductory workshops, the Modern Meta-Analysis Research Institute (MMARI) is targeted at early-career researchers with no previous experience with meta-analysis. Unlike other introductory workshops, the content of this Research Institute will focus on providing the data analysis skills in R to implement best-practice statistical methods.  At the conclusion of the workshop, participants should be able to: use R for meta-analysis; understand differences between effect sizes and compute effect sizes from the most common types of data reported in studies; specify an appropriate meta-analysis model; estimate and report both an average effect size and the extent of variation in effect sizes; explore and interpret heterogeneity of effect sizes using meta-regression models, and conduct appropriate publication bias analyses and interpret the effect of possible bias on findings.