Dear Enzo,
For two conditions within-subjects you should use the paired t-test and for more than two conditions the flexible factorial design, with factors 'subject', independence 'yes' and 'condition', independence 'no'. An F test over conditions is analogous to a traditional repeated-measures ANOVA taking into account non-spericity.
Best
Martin
Martin Dietz
[log in to unmask]
On Feb 15, 2012, at 2:30 PM, Enzo Tagliazucchi wrote:
> Hi all,
>
> I have two groups of subjects which I want to compare and some of them have repeated measurements. I am wondering which design I could use, since a simple t-test would be innapropiate (I'm looking for something like a repeated measures ANOVA). In the 2nd level specification menu of SPM the "independence" option (setted to "no") seems correct, but still I am not sure how this is modeled (it seems to me that since I am not specifying exactly which ones are the repeated measurements information is lacking).
>
> Thanks for any insights,
>
> Enzo
>
>
|