Meta-analysis pools data across studies addressing the same question, weighting each by precision to produce a combined estimate with narrower confidence intervals than any individual study.
It requires that the pooled studies be sufficiently similar.
Alternative Names:
Pooled Analysis, Quantitative Synthesis
Why it Matters?
Study heterogeneity is the methodological question that determines whether a meta-analysis is meaningful, since pooling studies with different exposure definitions, populations, or outcome measures produces a number that describes nothing real. Publication bias is the other vulnerability, because studies finding no effect are published less often, which skews any pooled estimate upward. Funnel plot analysis tests for that skew.
Frequently Confused with
Related terms
Frequently asked questions
What is heterogeneity?
What is publication bias?
