Thank you Helmut for your helpful comment!
This sounds like a very valid point (additional to Jesper’s arguments)!
Do you know, by any chance, if this is somewhere officially mentioned as a potential confound when using unwarp and the rp files together?
Von: MRI More [mailto:[log in to unmask]]
Gesendet: Freitag, 28. Juni 2019 13:37
An: [log in to unmask]; Manuel Kuhn <[log in to unmask]>
Betreff: Re: Realign & Unwarp - not including motion parameters as nuisance covariates
The rp files are also meant to account for interpolation errors that follow from realignment, at least this is what some standard textbooks tell you, and these would also occur in the case of realign & unwarp. Of course, these interpolation errors could be marginal, but I'm not aware of any more recent literature.
In any case, if one were to include the rp files there is an issue insofar as the rp files are generated after the realignment part, and they are not updated accordingly to reflect the adjusted parameters estimated during the unwarp part. I had mentioned this on the list a few years ago https://www.jiscmail.ac.uk/cgi-bin/webadmin?A2=SPM;c313ed02.1507 , but it didn't attract attention back then (might have been the usual issue with messages at the end of a month) and I didn't care much either, as it was just something that I had observed when trying out the module without further plans to use it.