Meta-analysis as a Method

Pooling effects across studies

Meta-analysis statistically combines effect sizes from multiple independent studies to produce a single pooled estimate. Each study is typically weighted by the inverse of its variance, so larger and more precise studies exert greater influence on the final result. As the quantitative core of a systematic review, meta-analysis converts a scattered literature into a single, interpretable summary. Assessing heterogeneity and publication bias is essential for drawing trustworthy conclusions.

What Is Meta-Analysis?

Meta-analysis is a statistical method that quantitatively synthesizes the findings of multiple studies examining the same research question. Individual studies may fail to reach definitive conclusions due to small sample sizes or methodological limitations; meta-analysis addresses this by combining studies through a common effect-size metric, yielding greater statistical power and narrower confidence intervals. The method was conceptualized by psychologist Gene Glass in 1976 and has since become a standard tool in medicine, education, psychology, and the social sciences, among other fields.

How It Works: Key Steps and Models

A meta-analysis typically follows these steps: (1) systematic literature search and definition of inclusion criteria; (2) extraction or conversion of effect sizes (Cohen's d, Hedges' g, odds ratio, r, etc.) from selected studies; (3) weighting each study by the inverse of its variance; (4) computing the pooled estimate using either a fixed-effect or random-effects model. The fixed-effect model assumes all studies share a single true effect, while the random-effects model allows for variation in true effects across studies and is generally preferred when the literature is heterogeneous.

A Concrete Application: Example Scenario

Suppose a researcher is examining the effect of cognitive-behavioral therapy (CBT) on depression. Twenty different randomized controlled trials may be available, each with varying sample sizes and effect sizes. Meta-analysis combines these studies into a weighted average, with larger and more precise studies carrying greater weight. The researcher ultimately obtains a single mean effect size and confidence interval showing that CBT significantly reduces depression relative to a control group. The I² statistic reveals the degree of heterogeneity across studies, informing judgments about the generalizability of the finding.

Common Pitfalls and Good Practice

The most frequently criticized problem in meta-analysis is excessive heterogeneity, also known as the apples-and-oranges problem: pooling studies that differ substantially in methods and context can produce misleading results. Publication bias is another serious threat; if only positive findings are published, the pooled estimate may be inflated. Funnel plots and Egger's test are used to detect such bias. Additionally, the quality of included studies should be evaluated using tools such as GRADE, and methodological limitations of primary studies should be reflected in overall conclusions. A well-conducted meta-analysis must be reported in accordance with the PRISMA guidelines and maintain transparency at every step.

Key terms

Effect Size: A quantitative value expressing the magnitude of a difference or relationship in a standardized metric.
Heterogeneity: Statistical variability reflecting true differences in effects across studies; measured by I².
Random-Effects Model: A pooling approach that allows for genuine variation in true effects across studies.
Publication Bias: Systematic distortion of pooled estimates caused by the selective publication of positive findings.
Inverse-Variance Weighting: Standard meta-analytic weighting where each study receives weight proportional to its estimation precision.