I've got two MaxDiffs in the same survey, measuring two different sets of features.
One that is a simple Most/Least Appealing, with one set of features.
The other one is about another set of features, but the metric we ask changes per type of respondent. For some the metric is "Most/Least Likely to Get me to Purchase", for others it's different such as "Most/Least Likely to Get me to Use the Product"
I assume for the latter I need to establish a covariate since the metric is the same between the groups, but I'm also wondering if I need to use a covariate for the simpler metric (the appeal one) even though , lest it look strange when one set of MaxDiff utility scores have lots of differentiation between the group, while the other set of MaxDiff utility scores has little differentiation