Jeff Sadino wrote:
Hello, I am running the mri_glmfit and would like to doublecheck my contrast file. This is my fsgd file:
MeasurementName thickness Class ctrl Class cd Class nc Variables age ed Input 040002_S04 nc 42.62 18 Input 040005_HG01 nc 45.26 15 Input 040008_S01 nc 26.92 16 Input 050015_S01 nc 46.65 12 .. .. ..
with values for cd and ctrl as well. I want to see if there are differences in thickness between my control group and my cognitive deficit group after taking into account for age and education. This is my matrix file: 1 -1 0 0 0 0 0 0 0 for ctrl, cd, nc, ctrl age, cd age, nc age, ctrl ed, cd ed, nc ed
Will this matrix file work for my hypothesis? It seems like if the last 6 are zeros, then it will not be taking age and education into account?
Yes, age and education are taken into account when the model is fit. Giving them 0s in the contrast implicitly means that they are nuisance and should be ignored. Also note that this contrast will test for a group difference at age=0 and education=0, which may or may not be what you want. If you want to test at the mean of your sample, then demean age and education using the entire sample mean (do not demean on a group-by-group basis).
And then also, since I have 90 subjects in all, will I need to correct for multiple comparisons?
The number of subjects does not matter for mult comp cor, but you will need to correct for spatial comparisons.
Thank you in advance, Jeff Sadino
Access your email online and on the go with Windows Live Hotmail. Sign up today. http://windowslive.com/online/hotmail?ocid=TXT_TAGLM_WL_HM_AE_Access_022009
Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer