comparison between PLS and SPM
as22kk
Posted on 08/14/08 10:41:04
Number of posts: 60
Hello randy again,
Well, what you said is not working for the region(which we are mostly interested) with highest bootstrap ratio.you can find the response plot for this region at: ftp://ftp.physiol.umu.se/fysiologi/out/Alireza/
To my opinion is is well enough to be capture by canonical HRF???
/Alireza
Untitled Post
rmcintosh
Posted on 08/16/08 11:41:29
Number of posts: 394
I can't access the FTP directory you sent as it requires a user name and password.
In any case, my guess as to the source of the differences by SPM and PLS may have to do with the hierarchical testing in SPM vs the resampling framework for PLS, which then also interact with multivariate vs univariate assessments. If you look at your region of interest in individual SPM's, do any/all of the subjects show an effect there?
Untitled Post
rmcintosh
Posted on 08/16/08 11:45:09
Number of posts: 394
Sorry, I accessed the image of the rf plot for the region. The effect should have been captured by a canonical HRF at least based on the average plot. I would still check to see whether the region shows up at the individual SPM level.