Andreas, thanks for your response.
So is this why I'm not seeing mainly any activation for the t-contrast at
0.05 for the F-test. Both my p-values are thresholded at 0.05 for both the
t-contrast and F-test. So you are saying that my t-contrast is at 0.025 in
the positive direction for the t-contrast and 0.05 (positive and negative)
for my F-test?
On Sun, 29 Oct 2006 19:05:22 +0100, Andreas Bartsch
<[log in to unmask]> wrote:
>Hi Micheal,
>
>and be aware that you may need to double your p-value for the F-test in
order to get the same probability under the tails, i.e. for your positive
and negative t-contrasts thresholded at p=0.05 you may need to use a
p=0.10 for the F-contrast to capture all of the latter.
>Cheers-
>Andreas
>
>Von: FSL - FMRIB's Software Library im Auftrag von Tim Behrens
>Gesendet: So 29.10.2006 17:41
>An: [log in to unmask]
>Betreff: Re: [FSL] couple of other things about f-test.
>Hi michael
>
>On 29 Oct 2006, at 15:57, Michael wrote:
>But I guess in the case w/ FSL, when setting up the contrast, it will
>always just show results for positive activation? Is there a way to
>indicate wanting to detect both positive and negative activation in the
>contrast? Or do we always have to include the F-stat for that?
>
>
>Just do a second contrast that is -1, this will give you a t-stat for the
>negative effect.
>
Would the setup be like this:
EV1 EV2 EV3 F1 F2 F3
OC1 * 1 0 0 * 0 0
OC2 * 0 1 0 0 * 0
OC3 * 0 0 1 0 0 *
OC4 * -1 0 -1 * 0 0
OC5 * 0 -1 0 0 * 0
OC6 * 0 0 -1 0 0 *
>Tim
|