Dear Jin,
An F-contrast defines a subspace of the design matrix (i.e.
a collection of t-contrasts) that may contain a significant
effect. The ensuing F statistic, if significant, allows you
to say that some linear combination of effects in this subspace
is significant. The F-contrast can be specified in a number of
ways. In the example below it is specified by the indices of
columns in the design matrix that are uninteresting (iX0). The
utility spm_FcUtil appends the contrast array so that you can
examine the associated SPM{F}.
see:
Büchel C Wise RJS Mummery CJ Poline J-B Friston KJ
Nonlinear regression in parametric activation studies
NeuroImage 1996;4:60-66
for an application of SPM{F}
I hope this helps - Karl
At 16:11 03/01/2003 -0500, you wrote:
>Dear Karl,
>
>I understand most parts of the spm_batch code. However, I couldn't figure
>out the F-contrast part. Could you please explain more?
>
>Best,
>Jin
>
>
>% F-contrasts
>%---------------------------------------------------------------------------
>%iX0 = [3:12];
>%cname = 'Visual';
>%SPM.xCon(end + 1) = spm_FcUtil('Set',cname,'F','iX0',iX0,SPM.xX.xKXs);
>
>
>
>
|