An introduction to meta-analysis with applications in R

Meta-analysis is the process of combining the estimates (effects) from several different studies to obtain a single estimated effect. By pooling these effects, we can increase power and explore new research questions. While conducting a meta-analysis in R is not a difficult thing to do, it is true that many published meta-analyses are not done well. This is because there is much to think about, including which model to use, the importance and acceptance of heterogeneity and what that means for interpretation, the potential for publication bias and so on. In this session I will introduce meta-analysis and highlight a few things that need to be considered to ensure that the analysis can be trusted. I will show how simple it is to carry out a meta-analysis in R on some real data sets.

Bio

Luke is the Associate Head of School, School of Engineering and Mathematical Science having previously been the Head of the Department of Mathematics and Statistics for nearly six years. He is also Director of La Trobe’s Statistical Consulting Platform and dedicates at least half of his research time to collaborative research spanning several disciplines. Luke’s interests in statistical methodology research include meta-analysis, robust statistics and visualisation of high-dimensional data. He is, and has previously been, a member of grant review panels for the National Health and Medical Research Council.

Zoom link – link will be provided by a calendar invite