Preference taks

[ferponcem]

It's a bit puzzling that we don't have a proper description for Preference as a stand alone task in the documentation, but we have contrast and tags for it (see here), although I am not sure all of them are well defined from utils_contrast.
Similarly, we have contrast maps for this task in derivatives but not in the smooth version.

This is just one protocol btw, so I am wondering whether we should...
a) treat this protocol as 5 different tasks (Preference, PreferenceFaces, PreferenceFood, PreferenceHouses and PreferencePaintings) and add a proper description for each; or
b) leave it as just four and remove the Preference contrasts from all_contrasts

[bthirion]

The point is that in Preference we put the fixed-effects analysis to get the common effect of expressing a preference or uncertainty thereof. Maybe we should call it PreferenceFFX instead ?

[ferponcem]

Oh, I got confused because Preference has some contrasts like preference_linear, preference_constant and preference_quadratic that I thought to be standalone contrasts, but are they only the joint effect for faces_linear + food_linear + ...??

[bthirion]

See https://github.com/individual-brain-charting/public_analysis_code/blob/master/scripts/script_preferences.py

[ferponcem]

Oh, I understand better now, thanks !
In this case I can just add a Preference task description specifying those are the fixed effects of the 4 previous tasks. This way we don't have naming to adjust, just in the documentation.

[bthirion]

I've updated the dataset, should be all right now...

[ferponcem]

Thanks! The only thing I see is that subjects 01 to 09 have the contrast named as Preferencelinear.nii.gz, while subjects 11 to 15 have it as preference_linear.nii.gz.
Probably harmless...