Alex,

The answer to the second question is yes (that group contrasts are run level_1 - level_2). When the qdec Analysis button is pressed, each contrast matrix is printed to the terminal, something like this: . . . Diff-Demented-Nondemented-Intercept-thickness ----------------------- Does the average thickness differ between Demented and Nondemented? 1.000 -1.000; . . . so you can confirm which is being subtracted.

Per your first question, in a future version of qdec, the form of the question will include the extension ", regardless of <continuous factor>", to help clarify the meaning. For example, your example question would read:

"Does the average thickness differ between grp1 and grp2, regardless of age?"

Nick

On Tue, 2007-10-30 at 20:01 -0400, Doug Greve wrote:

The answer to the first question is yes, it will covary (or "regress out") for the continuous variable. For question #2, I also believe that the answer is yes, though Nick may be able to say definitively.

doug

Alexander Fornito wrote:

...

Hi all, If i run a model in qdec that inludes 1 discrete and 1 continuous factor, does the output for the group contrast covary for the continuous variable? For example, if I want to compare two groups, and include group as my discrete factor and age as a continuous factor, will the output for the group contrast (Does the average thickness differ between grp1 and grp2?) covary for age? I've notice that running it without continuous factors leads to different results, so I just wanted to confirm this is the case.

Also, just wanted to check that group contrasts are run as level_1 - level_2, so that positive values (i.e., red-yellow) indicate level_1 > level_2.

Thanks for your help, Alex _______________________________________________ Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer