Skip to content
Omnikalc

Meta-analysis calculator

Enter the effect size and the standard error of three studies to get the pooled estimate with fixed and random effects, their confidence intervals, the heterogeneity statistics Q and I² and the between-study variance.

Updated on
Your inputs

Value out of range

Value out of range

Value out of range

Value out of range

Value out of range

Value out of range

Results

Fixed-effect pooled estimate

0.4411

Random-effects pooled estimate

0.4964

Fixed effect, lower 95 % limit
0.2291
Fixed effect, upper 95 % limit
0.6531
Random effects, lower 95 % limit
0.1106
Random effects, upper 95 % limit
0.8821
Cochran’s Q
5.993
p-value of Q (2 degrees of freedom)
0.0499
I² heterogeneity
66.6 %
Substantial heterogeneity: prefer random effects
Between-study variance τ²
0.0768

The meta-analysis calculator pools the results of three studies with fixed and random effects, and gives the confidence intervals, Cochran’s Q, I² and the between-study variance.

How the calculation works

  • Enter the effect size of each study.
  • Enter the standard error of each study.
  • The calculator gives the two pooled estimates with their 95 % intervals, Q, its p-value, I² and τ².

Example

Three studies find effects of 0.80, 0.20 and 0.55 with standard errors of 0.20, 0.15 and 0.25. The weights are 25, 44.4 and 16, so the fixed-effect estimate is 0.4411. Q = 5.993 with 2 degrees of freedom, so I² = 66.6 %, a substantial heterogeneity. The between-study variance is 0.0768 and the random-effects estimate is 0.4964, with a wider interval from 0.1106 to 0.8821.

Why it matters

The calculator helps students in epidemiology and evidence-based medicine, reviewers who check a forest plot, and researchers who need a quick pooled estimate of a few studies.

Good to know

A good meta-analysis rests on a systematic search and on studies that are similar enough to be combined. This calculator handles the arithmetic for three studies only; a real review needs software for more studies and for refinements such as the Hartung-Knapp adjustment, which gives better intervals with few studies. The DerSimonian-Laird estimator used here tends to underestimate heterogeneity when there are few studies. I² describes the share of variation due to heterogeneity and not its size, so also look at τ² and at the prediction interval. Always examine the studies one by one before trusting a pooled number, and report how the studies were chosen.

The weighted average calculator shows the weighting idea, the Cochran Q calculator tests binary matched data, the confidence interval calculator brackets means and the Cohen’s d calculator gives standardized effects.

Frequently asked questions

What does a meta-analysis do?

It combines the results of several studies into one estimate. Each study is weighted by the inverse of its variance, so that precise studies count more. The effect can be a mean difference, a log odds ratio, a log risk ratio or any measure with a standard error.

Fixed or random effects?

Fixed effect assumes that all studies estimate one true effect. Random effects assume that the true effect varies from study to study and adds the between-study variance τ² to each study’s variance. When studies differ, random effects give wider intervals and more even weights.

Should I use odds ratios directly?

No. Enter the natural log of the odds ratio and the standard error of that log, then take the exponential of the results to get back to the ratio scale. The same applies to risk ratios and hazard ratios.

Was this calculator helpful?

Thanks for your feedback!

Results are estimates provided for information purposes. They do not replace professional advice.