Sensitivity Analysis: Leave-One-Out
This lesson is part of the course
In this lesson
You re-run the fixed-effect meta-analysis eight times, each time without one study, find the most influential study and check whether the conclusion survives.
Buy the course once for $9.99 to unlock every lesson and lab, the final exam and the certificate – lifetime access, no subscription.
Lessons in this course
- Effect Sizes: Mean Difference, Cohen's d and the Log Odds Ratio Locked
- Inverse-Variance Weights & the Fixed-Effect Estimate Locked
- Confidence Interval & z-Test of the Pooled Effect Locked
- Heterogeneity: Cochran's Q, I² and τ² (DerSimonian–Laird) Locked
- The Random-Effects Model and the Prediction Interval Locked
- The Forest Plot: Every Study and the Pooled Effect in One Chart Locked
- Publication Bias: Funnel Plot & Egger's Regression Test Locked
- Sensitivity Analysis: Leave-One-Out Locked