Meta-Analysis

SW865, Section 001

This is a doctoral-level method course focused on advanced methodologies in research synthesis. Upon completion, students will be able to conduct a systematic review and meta-analysis study of publishable quality. The first two weeks of the course will provide a comprehensive introduction to systematic review, including developing a research topic, selecting databases, crafting search keywords, and conducting literature search and data extraction in a rigorous and transparent (replicable) way. The rest of the semester will focus on methodologies of meta-analysis. We will start with calculating standardized mean difference (SMD) as well as converting other statistics into the SMD metric. We will then discuss pooling SMD effect sizes using fixed versus random effects models. Meta-regression analysis will also be introduced to facilitate moderator analyses. Advanced methods to address dependent effects, i.e., multi-level meta-analysis, generalized least squares, and robust variance estimation, will be introduced and practiced during the latter part of the semester. Additional topics including assessing publication bias, meta-analysis of correlations and meta-analyzing categorical effects will be discussed if time permits. The R Software will be used for all analyses in this course.