There are several places where there could be problems.
1. You did not transfer the cluster label files to the individual correctly. What was your command line? 2. The clusters are very small and do not transfer well back to the individual 3. You are not using the same model in SPSS as you are in mri_glmfit (this happens a lot)
One thing you can try is to generate the table without going back all the way to the individual. There should be a y.mgh file with a stack of all your subjects, you can run mri_segstats on that (with --avgwf output option).
On 11/16/2015 05:21 AM, Bittner, Nora wrote:
Dear FreeSurfer experts,
we calculated a GLM in qdec with our experimental variable (continuous), age and gender as covariates and LGI. We found significant clusters and corrected them for multiple comparisons with Monte Carlo Z Simulation. We found several significant clusters and looked at the correlation plots qdec displays automatically.
In qdec we created labels from this clusters and mapped these to all subjects of our table.dat.
Then we used mri_segstats to generate textfiles (.stats) listing our lgi values for each participant.
Afterwards we correlated the lgi values with our experimental variable to plot the correlation. We used partial correlations and corrected for gender and age in SPSS.
Sadly, we found no correlation between lgi values and our experimental variable anymore.
How can this statistically be explained?
Is there anything we did wrong?
Thank you in advance!
Best regards,
Nora and Christian
Forschungszentrum Juelich GmbH 52425 Juelich Sitz der Gesellschaft: Juelich Eingetragen im Handelsregister des Amtsgerichts Dueren Nr. HR B 3498 Vorsitzender des Aufsichtsrats: MinDir Dr. Karl Eugen Huthmacher Geschaeftsfuehrung: Prof. Dr.-Ing. Wolfgang Marquardt (Vorsitzender), Karsten Beneke (stellv. Vorsitzender), Prof. Dr.-Ing. Harald Bolt, Prof. Dr. Sebastian M. Schmidt
Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer