meta-analysis
meh-tah uh-NAL-ih-sis
ˌmɛtəəˈnæləsɪsListen
Pick from 16 AI voices across American and British accents, and adjust playback speed to hear each syllable clearly. Every voice is generated once per term and cached for instant replays.
Syllable breakdown
Dictionary meaning
a statistical method that combines the quantitative results of multiple independent studies addressing the same research question to produce a single, more precise and statistically powerful estimate of an effect size than any individual study could provide; meta-analysis is typically paired with systematic review — which identifies and qualitatively synthesises the available evidence — to produce the quantitative pooled estimate that places the synthesised evidence on a common metric; the core statistical procedure involves weighting each study's estimate by a function of its sample size and variance — usually the inverse of its variance — before calculating a weighted average; key analytical steps include testing for statistical heterogeneity — inconsistency in results across studies — and assessing publication bias — the tendency for studies with positive results to be published while null results remain unpublished; meta-analyses underlie clinical guidelines, regulatory decisions, and health technology assessments
Examples
- The meta-analysis of 47 randomised trials confirmed that the intervention reduced mortality by 22 percent across diverse settings.
- Funnel plot asymmetry in the meta-analysis suggested publication bias, with small negative trials apparently absent from the literature.