Effect sizes and meta-analyses: Tools for cumulative, robust experimental science
Email address: firstname.lastname@example.org
Website teacher: http://www.mpi.nl/people/bergmann-christina
Title of the course: Effect sizes and meta-analyses: Tools for cumulative, robust experimental science
Teacher: Christina Bergmann
Single studies have long been the norm in experimental sciences to establish “facts” about the world, with little regard for cumulative thinking and reproducibility of the reported effects. This is problematic because our statistical tools are never completely conclusive. One consequence is that results frequently cannot be replicated, either because the effect is smaller than reported or because it simply is not present in the general population. Cumulative science, i.e. considering multiple studies together to get a better idea of what might be true, is one answer to this problem.
This course will give an introduction to tools of cumulative science: effect sizes and meta-analytic methods, including how to determine sample sizes before running a study and making informed design choices. Attendees are expected to have basic knowledge of the R statistical programming language and of standard frequentist statistical tests (t-tests, correlations, linear models).
Why cumulative science matters: Reproducibility, replicability, and robust science.
Theoretical introduction to effect sizes and meta-analyses, and how to consider replicability from the start.
Practical introduction to conducting reproducible meta-analyses 1: Systematic literature review, computing effect sizes from different reported statistics.
introduction to conducting reproducible meta-analyses 2: Meta-analytic models,
interpreting the output, conducting moderator analyses.