Meta-analysis

22 & 23 novembre 2018, 09:00-18:00

Paul-Christian Bürkner, PhD, University of Münster

In this two-day workshop, we will teach the statistical basics of meta-analysis. At the first day, we will start with the very idea of meta-analysis, discuss the computation of effect sizes from information provided in the primary studies and fit simple fixed and random effects meta-analytic models. Moreover, publication bias and other common biases in meta-analysis will be discussed in detail. During the second day, we will show how to perform moderator analysis as well as multilevel meta-analysis if multiple effect sizes per study or sample are provided. The discussion of each topic will be followed by a hands-on exercise in the statistical language R. The goal is that at the end of the second day, participants can perform all analysis necessary for most common types of meta-analysis found in various fields of psychology.As the workshop will not provide a general introduction to R, participants are expected to have at least some basic knowledge of R. This includes reading in and simple transformations of data sets as well as fitting linear models. All necessary knowledge to perform Meta-Analysis in R will be taught during the workshop.

UNIL, Bâtiment Géopolis, salle 4314 (le 22.11), salle 2218 (le 23.11)

20