Even if you use the same subjects, you are performing three different analyses, so it is not surprising that you are getting different results. As to why they are wildly different, I don't know. I suspect that you need to demean your continuous variables.
On 2/10/2020 11:26 AM, Wang, Lily wrote:
Hi Freesurfer Team,
I made a qdec table with columns fsid, disease status (control and diseased), age, df, db, dtotal (which are three continuous psychological test scores). I performed one analysis in qdec with disease status as the discrete factor and each of the three test scores as the continuous factor with no nuisance factor chosen, for a total of three analyses. All three analyses ran successfully but when I choose "Does the average thickness differ between Control and Diseased?", which should ignore the test scores (right?), and applied an FDR = 0.05, I get wildly different results among the three analyses even though I believe I should get the same results since it is the same pool of subjects in all three analyses. Did I do something wrong?
Thank you, Lily
Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer