You can't just go and regress out age. You first have to make sure that there is not an interaction between your age and group. That is done with the first step that I described using the DODS model. Both of the contrasts that you have below are for DODS, but you would only use the 2nd one. I know that you are not interested in age, but you have to test for an interaction with age first. If there is an interaction, then your contrast of interest (the difference between the group means) does not make sense. The first contrast below does test for a difference in the groups at age=0. If there is an interaction between age and group, then the difference between the groups becomes dependent on age. Eg, you can get a different difference if you test at age=0 vs age=50. If there is no difference between the age slopes of the groups, then you re-run the analysis using a DOSS model and test with a contrast of [1 -1 0]. doug
Antonella Kis wrote: > Dear Doug, > > I'm very sorry to bother you so much but I am getting very confused > because on the GLM Group Analysis site I saw the following two situations: > > > Contrast 1 group.diff.mtx > > Null Hypothesis: is there a difference between the group intercepts? > Is there a difference between groups regressing out the effect of age? > 1 -1 0 0 > This is a t-test with Group1>Group2 being positive (red/yellow). > > > Contrast 2 group-x-age.mtx > > Null Hypothesis: is there a difference between the group age slopes? > Note: this is an interaction between group and age. Note: not possible > to test with DOSS. > 0 0 1 -1 > > So if I want to compare the two groups and consider age as a covariate (I am > not interested in the effect of age on my cortical thickness) then should I > go for the second case as you emailed me before? > I thought is the first case when we regress out the effect of age so we can > see the differences in the cortical thickness between the two groups while > age changes. > > I'm really sorry but I really > need your last advise. > > > Many, many thanks. > AK > > > > ------------------------------------------------------------------------ > *From:* Douglas N Greve <gr...@nmr.mgh.harvard.edu> > *To:* Antonella Kis <ator...@yahoo.com>; freesurfer > <freesurfer@nmr.mgh.harvard.edu> > *Sent:* Tuesday, September 27, 2011 12:44 PM > *Subject:* Re: [Freesurfer] Thank you > > If you have two groups and you want to test for the differences in the > groups after you account for age, you need to do this in two steps. > First, set up a DODS model (ie, each group gets an intercept and a > slope). Use this to test for a difference between the slopes of the > groups ([0 0 1 -1]). If there is a difference, then you'll need to talk > to a statistician to figure out what to do. If there is no difference, > then set up a new DOSS model (ie, each group gets its own intercept but > they share a slope), then test the difference between the groups with [1 > -1 0]. > > doug > > Antonella Kis wrote: > > Thanks Doug, I followed your advise and it worked. > > I would like to ask you one more favour please: if using GLM I want to > > compare two groups and consider age as a covariate, does this mean > > that we account for age or we are regressing out the effect of age? Is > > this right? In this case can I consider the contrast matrix to be [1 > > 0] since we need to add zero for the variable which are covariate? > > > > Thank you. > > AK > > > > ------------------------------------------------------------------------ > > *From:* Douglas N Greve <gr...@nmr.mgh.harvard.edu > <mailto:gr...@nmr.mgh.harvard.edu>> > > *To:* Antonella Kis <ator...@yahoo.com <mailto:ator...@yahoo.com>> > > *Cc:* "freesurfer@nmr.mgh.harvard.edu > <mailto:freesurfer@nmr.mgh.harvard.edu>" > <freesurfer@nmr.mgh.harvard.edu <mailto:freesurfer@nmr.mgh.harvard.edu>> > > *Sent:* Tuesday, September 27, 2011 11:17 AM > > *Subject:* Re: Clusters formation by FDR > > > > Why not run mri_glmfit-sim? It will give you the file that you want, > > and it will be easier than me explaining all the stuff you'd need to > > do to get it otherwise. > > doug > > > > Antonella Kis wrote: > > > Good morning Doug, > > > > > > > > > I will really appreciate if you can give me an advise on how I can > > get the mean average of the cortical thickness value for the FDR > > clusters so before Monte Carlo Simulation. > > > > > > Many thanks, > > > Antonella > > > > > > > > > > -- Douglas N. Greve, Ph.D. > > MGH-NMR Center > > gr...@nmr.mgh.harvard.edu <mailto:gr...@nmr.mgh.harvard.edu> > <mailto:gr...@nmr.mgh.harvard.edu <mailto:gr...@nmr.mgh.harvard.edu>> > > Phone Number: 617-724-2358 Fax: 617-726-7422 > > > > Bugs: surfer.nmr.mgh.harvard.edu/fswiki/BugReporting > > FileDrop: www.nmr.mgh.harvard.edu/facility/filedrop/index.html > > > > > > > > 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. > > > > > > > > -- > Douglas N. Greve, Ph.D. > MGH-NMR Center > gr...@nmr.mgh.harvard.edu <mailto:gr...@nmr.mgh.harvard.edu> > Phone Number: 617-724-2358 > Fax: 617-726-7422 > > Bugs: surfer.nmr.mgh.harvard.edu/fswiki/BugReporting > FileDrop: www.nmr.mgh.harvard.edu/facility/filedrop/index.html > > _______________________________________________ > Freesurfer mailing list > Freesurfer@nmr.mgh.harvard.edu <mailto:Freesurfer@nmr.mgh.harvard.edu> > https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer > > -- Douglas N. Greve, Ph.D. MGH-NMR Center gr...@nmr.mgh.harvard.edu Phone Number: 617-724-2358 Fax: 617-726-7422 Bugs: surfer.nmr.mgh.harvard.edu/fswiki/BugReporting FileDrop: www.nmr.mgh.harvard.edu/facility/filedrop/index.html _______________________________________________ Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer
