Then I'm not sure. If it was uncorrected then this could have something to do with the recently discussed problems of random field theory at low uncorrected thresholds. But with FWE it should all be fine. Perhaps some RFT experts could comment.

Vladimir

On Wed, Feb 3, 2016 at 9:20 PM, Keenan, Alexandra <[log in to unmask]> wrote:
Dear Vladimir,
I used 0.05 FWE.
Thanks,
Alexandra

On Wed, Feb 3, 2016 at 12:54 PM, Vladimir Litvak <[log in to unmask]> wrote:
Dear Alexandra,

Do you mean 0.05 FWE, or 0.05 uncorrected?

Vladimir

> On 3 Feb 2016, at 15:54, Alexandra Keenan <[log in to unmask]> wrote:
>
> Hi SPM experts,
>
> I have 2D time-frequency images averaged across EEG channels with 2 conditions per subject over 54 subjects. I have done a repeated measures ANOVA using T-contrasts with weights [1 -1] and [-1 1] and an F-contrast with a weight matrix [eye(2) ones(2,54)/54]. I kept the threshold at 0.05. There were no significant clusters in either of the T contrasts, but several in the F-contrast. I'm having trouble understanding how this can be.
> Any help would be greatly appreciated!
>
> Thank you,
> Alexandra