A procedure for statistically combining the results of many different research studies (comparing effect sizes) is called what?

• A. Meta-analysis
• B. Systematic review
• C. Narrative review
• D. Pooled analysis

Answer: (A)

The idea being tested is the statistical synthesis of results from multiple studies to arrive at a single overall estimate. This is meta-analysis. In practice, you take the effect size from each study (for example, an odds ratio, risk ratio, or mean difference), then combine these effects using a weighted approach that favors more precise studies (usually inverse-variance weighting). The result is a pooled effect that reflects what the body of research as a whole suggests, and it helps increase statistical power and clarify whether there is a consistent effect across studies.

Meta-analysis also involves looking at how much study results vary beyond chance, known as heterogeneity (for example, I-squared or Q statistics). If heterogeneity is present, a random-effects model may be used, acknowledging that the true effect could differ between studies; if not, a fixed-effects model may be appropriate, assuming a single common effect. Additionally, meta-analysis can include checks for publication bias (like funnel plots) to assess whether only certain results tend to get published.

A systematic review is the broader process of identifying, appraising, and summarizing all relevant studies on a topic with explicit methods; meta-analysis is the quantitative step that may be part of a systematic review. A narrative review is a qualitative, non-systematic summary, and pooled analysis is not the standard term used for this specific, published-data synthesis.