[Freesurfer] mri_glmfit: Hierarchical Regression?
Hello all, I am using mri_glmfit (DODS method) to run a CT analysis with 2 groups and 4 covariates and have a couple of questions regarding the statistical tests performed within the group analyses: 1. Can Freesurfer test the change in model fit or delta R-squared in a hierarchical regression? As far as I can tell, there is only simultaneous regression. Alternatively, can I change the Sum of Squares Type to Type 1 (I think default is Type 3) to simulate a hierarchical regression? 2. I think the default model in FreeSurfer accounts for all other main effects. But, does it also account for all 2 and 3 way interactions, whether or not I specify any? In other words, is the default a fully factorial model? Thank you very much for your time! -Elizabeth ___ Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer The information in this e-mail is intended only for the person to whom it is addressed. If you believe this e-mail was sent to you in error and the e-mail contains patient information, please contact the Partners Compliance HelpLine at http://www.partners.org/complianceline . If the e-mail was sent to you in error but does not contain patient information, please contact the sender and properly dispose of the e-mail.
Re: [Freesurfer] mri_glmfit: Hierarchical Regression?
Hello, It doesn't appear that my question posted to the forum. Is there anything I need to do in order for the question to post? Thanks, Elizabeth On Wed, Sep 28, 2016 at 6:03 PM, Elizabeth Bartlett < elizabeth.bartl...@stonybrook.edu> wrote: > Hello all, > > > I am using mri_glmfit (DODS method) to run a CT analysis with 2 groups and > 4 covariates and have a couple of questions regarding the statistical tests > performed within the group analyses: > > > > 1. Can Freesurfer test the change in model fit or delta R-squared in a > hierarchical regression? As far as I can tell, there is only simultaneous > regression. Alternatively, can I change the Sum of Squares Type to Type 1 > (I think default is Type 3) to simulate a hierarchical regression? > > 2. I think the default model in FreeSurfer accounts for all other main > effects. But, does it also account for all 2 and 3 way interactions, > whether or not I specify any? In other words, is the default a fully > factorial model? > > > Thank you very much for your time! > > -Elizabeth > ___ Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer The information in this e-mail is intended only for the person to whom it is addressed. If you believe this e-mail was sent to you in error and the e-mail contains patient information, please contact the Partners Compliance HelpLine at http://www.partners.org/complianceline . If the e-mail was sent to you in error but does not contain patient information, please contact the sender and properly dispose of the e-mail.
[Freesurfer] Adjusted or unadjusted CT values and interpretation of volume values in ocn.dat from group analysis
Hello Freesurfer experts, I have performed a group level analysis of volume and cortical thickness using mri_glmfit and performed multiple comparisons correction with mri_glmfit-sim. I used the contrast matrix 0 1 0 to analyze 1 group with 2 covariates, along the 1st covariate (education) with the 2nd covariate being age. I get 2 significant clusters (after mult. comp. corr.) for CT and 2 for volume. When I look at the values in ./cache.th30.abs.y.ocn.dat, the average CT values within my sample are 2.57 and 2.50 for the 2 regions. *My first question is: *are the CT values outputted to ocn.dat adjusted for my 2nd covariate (age) or are they unadjusted CT values within the sig. cluster? The volume values contained within the ocn.dat files are also in the same range as the CT values, with averages of 2.50 and 1.87 for left superior frontal and right inferior parietal volume, respectively. *My second question is:* What is the metric for volume reported in ocn.dat? These values seem too low to be total cluster volume for each subject expressed as mm^2. The volumes reported in ./cache.th30.abs.sig.cluster.summary are more like what we I expect for volume (283.24 mm^2). Could these values be the average volume surrounding each vertex in the inflated mesh (volume contained within the surrounding vertices)? Thank you, Elizabeth ___ Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer The information in this e-mail is intended only for the person to whom it is addressed. If you believe this e-mail was sent to you in error and the e-mail contains patient information, please contact the Partners Compliance HelpLine at http://www.partners.org/complianceline . If the e-mail was sent to you in error but does not contain patient information, please contact the sender and properly dispose of the e-mail.
