Which meta-analysis model assumes a single common effect size across all studies?

• A. Bayesian model
• B. Random effects model
• C. Fixed effects model
• D. Mixed effects model

Answer: (C)

In meta-analysis, the fixed effects approach assumes there is one true effect size shared by all studies. The differences you see in study results are attributed only to sampling error within each study, not to real differences in the underlying effect. Because of that, the overall effect is calculated as a weighted average of the study estimates, with weights usually based on each study’s variance (more precise studies pull the summary more). This means the combined result is intended to apply to the set of studies included, rather than to a broader universe.

If there’s genuine variation in true effects across studies (heterogeneity), this fixed view can be misleading, because it ignores between-study differences. In that case, a random effects model is used, which accounts for between-study variance and typically yields wider confidence intervals. Mixed effects and Bayesian frameworks are broader approaches that can implement either fixed or random assumptions depending on how they’re specified, but the defining property of a single common effect across studies points to the fixed effects model.