FREE TOOL
Compute a standardized effect size from raw study data — dichotomous 2×2 event counts, or continuous means and standard deviations.
OR
1.714 [95% CI 0.895, 3.285]
Odds ratio & risk ratio are computed on the log scale using Woolf's method — SE{ln(OR)} = √(1/a + 1/b + 1/c + 1/d) — then exponentiated back for display. A 0.5 continuity correction is applied to all four cells automatically if any cell is zero.
Risk difference is computed directly on the linear scale as the difference in event proportions between arms.
Hedges' g is Cohen's d with a small-sample bias correction factor J = 1 − 3/(4·df−1), where df = n₁+n₂−2 — the standardized mean difference convention used in Cochrane reviews.
FAQ
Hedges' g applies a small-sample bias correction to Cohen's d, making it slightly more conservative for smaller studies — it's the standard used in Cochrane-style meta-analyses.
Odds ratio is common in case-control studies and logistic regression; risk ratio is more directly interpretable for cohort/RCT data; risk difference gives an absolute (not relative) measure of effect, useful for clinical decision-making.
When any cell in the 2x2 table is zero, the log odds ratio or log risk ratio is mathematically undefined. Adding 0.5 to all four cells (the Haldane-Anscombe correction) is the standard fix.
This computes one study's effect size at a time. To pool multiple studies with heterogeneity statistics and forest plots, use the I² calculator alongside this one, or run a full analysis in EvidenceFlow.
Extracting effect sizes for more than one study? EvidenceFlow auto-fills these fields from full text and pools them into a forest plot automatically.
Create free account