Evidence Synthesis 2: Statistics in Systematic Reviewing

Full course description

There are a lot of scientific publications. It is estimated that 1.8 million articles are published each year. Even in any chosen specific field tens of thousands of articles are published each year. For example, during the COVID-19 outbreak 23,500 articles were published on the topic in just the first wave. Any researcher or research-based professional is expected to synthesize the results of scientific studies for evidence-based decision making, regulatory approval or to identify the gaps in literature that need further research. Research synthesis and systematic reviewing are rapidly evolving academic fields using dedicated study designs, software, and statistical tools with applications in all research domains. In this semester, containing two skill trainings (in periods 4 and 5) and a project (in period 6), we will discuss the full scope of principles, concepts and methods of systematic literature reviewing, including meta-analysis (statistical pooling of outcomes of included component studies). You will also gain hands-on coding experience with the statistical programmes JASP and R. Having some experience with statistics or coding will thus help but is not a prerequisite. To facilitate the transparency requested of the modern scientist, you will be working in the Open Science Framework. The popular semester will teach you how to read and write academic papers. It is, as such, a good preparation for your capstone project and possibly later in your educational and academic career.

This second part of this semester is a skills course, which goes beyond the systematic review and give you the required statistical background to conduct a meta-analysis, a quantitative summary of all collected evidence. A meta-analysis synthesizes a body of research investigating a common research question. Outcomes from meta-analyses provide a more objective and transparent summary of a research area than traditional narrative reviews. Moreover, they are often used to support research grant applications, guide clinical practice, and direct health policy. An improved understanding this tool will not only help scientists to conduct their own meta-analyses but also improve their evaluation of published meta-analyses. You will be training in the statistical software R and run analyses for the identification of publication bias; (semi-)quantitative pooling of component study results (research synthesis, e.g., statistical pooling, best-evidence synthesis); assessment and exploration of heterogeneity of study results (e.g., outlier analysis, cumulative meta-analysis, meta-regression analysis); levels of evidence and interpretation of meta-analytic results; and computer software for meta-analysis.