Fixed & random-effects meta-analysis
Set up, run, and review this analysis in Lifescio.
Supported analysis
Supply an effect estimate and standard error per study. The app calculates inverse-variance common-effect and REML random-effects summaries, forest and funnel plots, a prediction interval and leave-one-out results. Enter ratio effects on the log scale.
Supported scope: Studies must measure compatible effects. Correlated estimates from the same participants need a multivariate or dependent-effects approach. Few studies make heterogeneity and prediction intervals unstable; numerical pooling cannot repair biased source studies.
Prepare data and configure the analysis
Enter one independent study per row: Effect, strictly positive Standard error, and any additional example moderator column. The pooling workflow uses the first two columns. Select difference for an additive effect or log-ratio when the supplied effects and standard errors are already on the natural-log scale.
Selecting log-ratio does not log-transform raw input ratios. Convert ratios and their uncertainty consistently before importing. At least three studies are needed by this implementation. Use a separate meta-regression workflow when moderators are part of the model.
Input reference (3 fields)
Enter your data and prespecified choices. Ranges include both limits. For table fields, use cell editing, Paste data, or Import CSV / TSV. Review whether the first imported row contains headings and map the source columns before applying the import. Empty numeric cells are not treated as zero. Open file is for an exported Lifescio analysis JSON file, not a spreadsheet.
| Input | Format | Allowed range | Guidance |
|---|---|---|---|
| Study estimates | Table | — | Use log effects for ratios. Each row is an independent study; moderator is used only in meta-regression. Columns in order: Effect; Standard error; Moderator. |
| Effect scale | Selection | — | Choices: difference, log-ratio. |
| Significance level | Number | 0.001–0.2 | — |
Run and review
- Give the analysis a descriptive name and review every example value. Record units, data preparation, and planning assumptions in Reproducibility & notes. For simulations, record the seed and replication or sampling settings.
- Choose Run analysis. Resolve input errors before interpreting output. In the connected app, a run uses a saved input revision; local preview runs need the page to stay open.
- Review Diagnostics for method limitations, warnings, and available numerical checks. A completed run means computation finished; it does not establish that the assumptions fit your study.
- Use the Result table selector to inspect each output. Editing inputs does not recalculate existing results: an earlier-input notice means you need to run again to obtain matching results.
Understand the outputs
Pooled effects remains on the analysis scale for both common and random effects. With log-ratio selected, Pooled ratios and the forest plot show exponentiated results. Heterogeneity includes tau squared, I squared, and prediction limits on the analysis scale.
Random-effects estimation uses restricted maximum likelihood and modified Knapp–Hartung t intervals. Study effects includes weights; Leave-one-out appears from four studies onward. The funnel plot also uses the analysis scale. Keep those scale distinctions clear when exporting tables.
Troubleshooting and common mistakes
A nonpositive standard error is invalid; do not supply variances or confidence-limit widths in that column. Extremely large ratios often indicate raw ratios were entered where log ratios were required.
If heterogeneity or leave-one-out results are unstable, investigate extraction, comparability, and the number of studies. A warning about few studies is a limitation of inference, not an instruction to add ineligible studies. Shared-control or repeated estimates require a dependence-aware method outside this simple workflow.
After correcting inputs, run again and check that the results belong to the new input snapshot. Increasing simulation draws improves numerical precision only; it does not repair invalid data or an unsuitable model.
Save, compare, and export
In the connected app, Save analysis retains the inputs in your account. Run history lets you inspect earlier runs, each with the input revision it used. In local preview, Save analysis downloads an input file instead; download results before leaving the page.
Use Result + inputs for a JSON archive containing the completed result and its inputs, Download CSV for the selected result table, PDF report for a report, and a chart's SVG button for its figure. Table displays round to three decimals; CSV and JSON exports retain calculation precision. Use those values when assessing a result near a decision boundary.
To reopen an archive, select the matching analysis method and choose Open file. Imported inputs start a new draft; a result archive can also display its retained result. It does not attach itself to an existing account analysis or recompute the result. Save or run the imported draft when you want a new account record.