Sharmii, You should also remove the covariate as covariates aren't valid in repeated-measure GLMs, only linear mixed-effects models. Best Regards, Donald McLaren, PhD On Tue, Jul 19, 2016 at 5:24 AM, Shane Mckie <[log in to unmask]> wrote: > Hi Sharmii > > Have you set the variance for the subject factor to equal? > > bw > > Shane > > Shane McKie > Neuroscience and Psychiatry Unit > Stopford Building > University of Manchester and Manchester Academic Health Sciences Centre > 0161 275 1740 > > -----Original Message----- > From: SPM (Statistical Parametric Mapping) [mailto:[log in to unmask]] On > Behalf Of Sharmili Edwin > Sent: 19 July 2016 08:26 > To: [log in to unmask] > Subject: [SPM] Flexible Factorial design taking very long > > Dear SPM experts, > > I'm running a seed based connectivity analysis with resting state data and > a pharmacological intervention. > > I have thirteen subjects with four intervention doses and seven timepoints > (4x7). > > So for each subject I have 28 conditions. So far I have used a full > factorial design. To consider Subject as a factor, I would like to run a > flexible factorial design. I'm also including one covariate. > > The fullfactorial model took only a few minutes on a server using matlab > 2014b. > > However, the flexible factorial design is running more than a week. It is > still checking for temporal non-sphericity. > I'm not getting any error message. At some point matlab crashes. > > Do you know what might be wrong or what i could improve? > > sharmili >