multiple comparisons correction
Statistical procedures applied to control false positive rates when performing numerous simultaneous statistical tests, essential in neuroimaging mass univariate analyses.
Full Definition
Multiple comparisons correction refers to statistical methods used to adjust significance thresholds when performing many simultaneous statistical tests, preventing inflated Type I error rates. In neuroimaging, where thousands or millions of voxels are tested independently, the probability of false positive findings increases dramatically without appropriate correction. Common correction methods include Bonferroni correction (most conservative), false discovery rate (FDR) control, and family-wise error rate (FWER) approaches such as random field theory. The choice of correction method involves balancing protection against false positives with maintaining statistical power to detect genuine effects. Cluster-based corrections and small volume corrections for regions of interest represent additional strategies for managing multiple comparisons.
Usage
Usage note: Essential to specify which correction method was applied in neuroimaging studies.
In Context
- "Results survived multiple comparisons correction using FDR at q < 0.05." — Neuroimaging statistics section
- "Cluster-level inference was performed to control for multiple comparisons across the whole brain." — fMRI analysis methodology