help > RE: Why are task conditions recommended to be used as confounds in denoising step?
Jun 22, 2022  10:06 PM | Alfonso Nieto-Castanon - Boston University
RE: Why are task conditions recommended to be used as confounds in denoising step?
Dear Jette,

The short answer is that the gPPI first-level analysis model explicitly includes the "task effects" as control covariates (the "psychological term"), so it does not matter whether you remove those effects or not during denoising. See the following thread for a few more details https://www.nitrc.org/forum/message.php?msg_id=26650

Best
Alfonso

Originally posted by Jette de Vos:
Dear,

We want to perform a gPPI analysi. As a default option the task conditions are entered as confounds in the denoising step, but these are the effects we are interested in, so we were wondering why it is recommended to regress them out as confounds?

If it does indeed make more sense to not include the task conditions as confounds, how do I prevent them from being used as confounds? I ran a batch script with no confounds specified, however, when loading the batch results in the GUI, the task conditions are still listed as confounds.

Best regards,
Jette

Threaded View

TitleAuthorDate
Jette de Vos Jun 21, 2022
RE: Why are task conditions recommended to be used as confounds in denoising step?
Alfonso Nieto-Castanon Jun 22, 2022