Re: [Freesurfer] selxavg3-sess: F-contrast but multiple dimension in sig

if you used the identity matrix then it is the first one On 02/08/2017 08:43 AM, Seung-Goo KIM wrote: > Thank you very much, Doug, > > One more question: so the sig have all T-test results of contrasts > like [1 0 0 … 0], [0 1 0 ... 0], …, [0 0 0 … 1], right? or [1 -1/(N-1) > -1/(N-1), …

Re: [Freesurfer] selxavg3-sess: F-contrast but multiple dimension in sig

Thank you very much, Doug, One more question: so the sig have all T-test results of contrasts like [1 0 0 … 0], [0 1 0 ... 0], …, [0 0 0 … 1], right? or [1 -1/(N-1) -1/(N-1), … -1/(N-1)] ? Best, -- Seung-Goo KIM > On 2017-02-06, at 18:54, Douglas N Greve wrote: >

Re: [Freesurfer] selxavg3-sess: F-contrast but multiple dimension in sig

yes, fsig is the F test. The sig has the result of all the individual t-tests On 02/06/2017 08:41 AM, Seung-Goo KIM wrote: > Dear surfers, > > I computed F-contrast using selxavg3-sess with an identity matrix as a > contrast matrix. My question is about the result file “sig.*”. As far > as I

[Freesurfer] selxavg3-sess: F-contrast but multiple dimension in sig

Dear surfers, I computed F-contrast using selxavg3-sess with an identity matrix as a contrast matrix. My question is about the result file “sig.*”. As far as I understand, the “sig.*” file contains signed-minnus-log-p-values (like, -log10(p)*Sign(t)) for a certain contrast. So I expected the