Hi Sandra,

That depends on what you'd like to test. By "combining the several runs" do you mean really combining them to see if they jointly are associated with the behavioural measure (one measure per session), or do you want to test if differences among them (e.g., after some intervention between the runs) is related to behaviour? Or do you have a behavioural measure per run, and would like to combine these inferences?

Merging the outputs of stage 2 per participant would only allow inference for that participant (N subjects would mean N hypotheses, one for each subject). It doesn't seem usual, although not impossible.

All the best,

Anderson

On Wed, 25 Jul 2018 at 14:27, Sandra Thijssen <[log in to unmask]> wrote:

Hi experts,

I have seen a few older threads on this subject, but just wanted to make sure the advice is still the same.

I am using resting-state data that was collected over multiple runs within the same session. I am interested in examining associations between a continuous behavioral measure and ROI functional connectivity. I was thinking of using dual regression for this purpose and was wondering how to go about combining the several runs. If I understand correctly, I should use dual regression without randomise on all runs for all participants separately. I should then combine all dual regression stage 2 outputs from the same participant using fslmerge and only then proceed to perform randomise.

Could you let me know if this sounds OK?

Thank you in advance!

Kind regards,
Sandra
########################################################################

To unsubscribe from the FSL list, click the following link:
https://www.jiscmail.ac.uk/cgi-bin/webadmin?SUBED1=FSL&A=1