Leave-one-study-out influence audit
Recompute a common-effect inverse-variance estimate after omitting each study.
Your entries stay in this page’s memory and are not submitted to us. Downloads are generated locally. Reloading or leaving the page clears your entries. Do not enter patient information.
Use 3–100 independent study estimates on one compatible analysis scale, with positive SEs. Log-transform ratio measures before entry.
The starting values are hypothetical. Results update as you edit. Use aggregate or synthetic inputs only.
Calculation results
- Full common-effect estimate
- 0.15081967
- Omission estimates in row order
- 0.08000000, 0.1076923, 0.2400000
- Largest absolute shift
- 0.089180328
- Most influential row
- 3
How to interpret the output
The pooled estimate weights each study by inverse variance; omission results assess influence under that common-effect model. They do not justify removing a study because its result is inconvenient. Shared participants, correlated estimates and between-study heterogeneity require other methods.
Method referenceSource checked 2026-11-06. Read the reproducible calculator validation report.