Loading Events

Abstract

Meta-Analyses are very popular in management research.  Such studies tend to be highly cited (Judge, Cable, Colbert, & Rynes, 2007); and play an important role in theory testing and refinement; evidence-based practice, and in turn policy-making (Banks, Kepes, & Banks, 2012). Given this popularity and ubiquity of Meta-Analyses in our scientific landscape, generating precise estimates of effects is necessary. Meta-Analyses rely upon an assumption of independence of primary effect sizes, which could be routinely violated in practice. Here I will discuss the implications of violating the independence assumption in meta-analysis and demonstrate how meta-analysis could be cast as a multilevel, variance known (Vknown) model to account for such dependency in primary studies’ effect sizes. I will walk us through a newly developed shiny app that helps cast a meta as a multi-level model when such dependencies exist in primary studies. This talk is based on the following article: Gooty, J., Banks., G.C., Loignon, A., Tonidandel, S., & Williams, C.W. (2021). Meta-analyses as a Multi-level Model. Organizational Research Methods, 24(2), 389-411.

*Multi-level Meta Shiny App: https://orgscience.uncc.edu/about-us/resources

Share This Story, Choose Your Platform!