If this refers to the standard error, shouldnt it be without the "head"?! Otherwise this is not consistent with the previous naming scheme

Shouldnt the "yerr" be the std_error calculated before?

And could someone explain the difference between the Standard error and the Standard Error of beta estimates...?

Shouldn't the main diagonal be all 1's, as the pattern similarity between the same conditions is perfect?

FYI: The code below "file_list[:20]" yields only the first 10 subjects, as there are is also a single .html file in the folder "fmriprep".

Didn't the procedure find 3 spikes? And wasn't a threshold of 2.5 Standard deviations used, not 3, as described above?

Might also be slightly confusing, as "video_computation_c0" appears to be the first regressor in the DM