How to do a meta-analysis
This page is about conducting a meta-analysis for a systematic review. For a checklist on writing a lab report based on an experiment or other original research efforts, see Checklist — Lab report write-up in Mike’s Genetics and Genomics Workbook.
Meta-analysis is a statistical approach to a systematic review — by combining data from many primary literature studies with similar design on the same question (outcome), the goal is to produce a single, pooled estimate of the effect, which increases statistical power and precision. As a crucial part of scientific reproducibility, meta-analysis is an important tool to help researchers move past the claims of a single study to see if results hold up on comparison from many studies. There are plenty of resources about and how to conduct a meta-analysis, many listed in Chapter 20.15 of Mike’s Biostatistics Book. Here, I provide links to three checklists that may help students conduct their first meta-analysis.
Although the assumption is the project is a meta-analysis, the writeup and video abstract guidelines apply to other kinds of biostatistics projects including original experimental work conducted by students in order to meet requirements of our biostatistics course.
Checklist — Basic meta-analysis