External Email - Use Caution
Hi doug, thanks for the quick response! I‘m not quite sure if I‘m getting the second part right- could you please elaborate a bit what you mean by „When there is only one numerator DOF, then the sig map is signed by the sign of the contrast (but the p-value is still based on an unsigned test).“ ?
It appears counter intuitive to me to visualize the results of a correlation analysis in a F map...
As I am mainly interested in the results if a (partial) correlation, I was wondering if there is a way to create a correlation map (e.g. based on the pcc.mgh overlay) that is masked by clusters surviving multiple comparisons corrections? (Monte carlo/ permutation)
Thanks again, Dorothee
On 3. May 2020, at 00:36, Douglas N. Greve dgreve@mgh.harvard.edu wrote:
A t-test is just F test with one DOF in the numerator -- they come out with the exact same p-value (assuming an unsigned t-test). When there is only one numerator DOF, then the sig map is signed by the sign of the contrast (but the p-value is still based on an unsigned test).
On 5/2/2020 5:32 PM, s7675854@stud.uni-frankfurt.de wrote: External Email - Use Caution
Dear Freesurfer experts,
I have a beginner question regarding fsgd and mtx files. For example, if I have one group and am interested in the correlation between some test score and cortical thickness/surface area, controlling for age and intracranial volume.
I would have 1 group and 3 covariates, e.g.
class 1
score age icv
subj1 1 55 20 3333
subj2 1 43 24 3333
Now if I am interested in the correlation between the test score and an anatomical parameter, I would go with the design matrix:
0 1 0 0 (doss). My question is, why does this calculate an F test, if I am only interested in one group? (I would understand a t-test with test-score slope >0, but where does the F come from?)
I am drawing a mental blank here and would appreciate your help!
Thanks a lot,
Dorothee
Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer
Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer