A quantitative synthesis method that combines effect estimates and their uncertainties from multiple empirical studies on a common question to produce aggregated estimates and to assess between‑study heterogeneity and related inference in political-science research.

Principle

Weighted aggregation (commonly inverse‑variance weighting) pools study estimates while modeling sampling error and between‑study variation; choices between fixed‑effect and random‑effects formulations formalize assumptions about study exchangeability.

Demonstration

Illustrative scenario: Situation — Ten studies estimate the same intervention’s effect on turnout but report different point estimates and standard errors. Recognition — Estimates vary beyond sampling error. Action — The analyst extracts effect sizes and variances, fits a random‑effects meta‑analytic model to obtain a pooled estimate and an estimate of between‑study variance, and reports heterogeneity statistics. Consequence — The synthesis yields an overall effect with quantified uncertainty and highlights whether effects differ systematically across studies.

Misapplication

Pooling studies that measure different constructs, combine incompatible populations, or ignore study quality and design differences; treating the pooled estimate as a single causal parameter without addressing study heterogeneity or publication selection.

Consequence

A well‑conducted meta‑analysis increases precision, clarifies the distribution of effects, and can reveal moderators or publication biases; an inappropriate meta‑analysis can produce misleading conclusions by aggregating incommensurate or biased evidence.

Reversal

If the underlying studies are not exchangeable (different estimands, populations, or systematic biases) or if selective publication skews the literature, a pooled estimate may not represent a meaningful population effect and pooled inference can be misleading.

Boundary

Clearly within: quantitative syntheses that extract comparable effect measures and variances from multiple empirical studies on the same question. Boundary case: combining studies with closely related but not identical outcomes where sensitivity analyses are required. Clearly outside: narrative literature reviews that do not quantify effects or heterogeneity.

Semantic Tension

Tension between increasing generalizability via aggregation and preserving contextual specificity—aggregation can mask conditional or population‑specific effects important for theory and policy.

Synthesis

Meta‑analysis formalizes cross‑study inference by weighting and modeling heterogeneity, but its substantive value depends on comparability and quality of constituent studies and on transparent modeling choices and sensitivity checks.