Modeling for Intergroup Comparisons of Imaging Data

摘要: Intergroup comparisons pose unique challenges in the analysis of functional imaging data. Imperfections intersubject stereotaxis can give rise to artifactual results and make it particularly important allow for differences task-related changes when formulating statistical models. Because intergroup generally involve inferences about populations from which subjects were drawn rather than particular themselves, must be treated as random fixed effects model. These requirements, combined with need adjust multiple spatial comparisons, result low power number each group is small. Functional studies identify between groups require many more other types careful advance planning maximize likelihood reaching meaningful conclusions.