New methods for the analysis of trial-to-trial variability in neuroimaging studies door