Yes. Certainly with subject-level covariates it is more appropriate to explicitly do the difference and then do a one-sample with covariates, because your covariates are not changing at the level of scan. And a paired T-test is actually statistically no different than a one sample T-test of differences, so it's definitely ok to do.
If you actually had per-scan covariates that differ within subject I'm not positive, but my impression is that SPM will not deal with covariates like these sensibly in repeated measures designs.
-Mike
> -----Original Message-----
> From: SPM (Statistical Parametric Mapping) [mailto:[log in to unmask]] On Behalf Of Karina Lopez
> Sent: Thursday, October 27, 2016 4:33 AM
> To: [log in to unmask]
> Subject: Re: [SPM] paired t test and nuisance variables
>
> Thank you!
>
> Just to understand better, a paired t test model in SPM does not allow the use of nuisance variables such age or sex? The approach
> you suggest me (difference images and then onesample t test) is statistically more appropriate?
>
> Thanks in advance!
>
> Karina
**********************************************************
Electronic Mail is not secure, may not be read every day, and should not be used for urgent or sensitive issues
|