This already exists but I must admit I am actually pro-considering as group_vars for PKNCAresults also start, end and perhaps in the future PPANMETH to differentiate any parameter that may not be calculated in the same way (e.g, different impute)

So in my opinion should be that group_vars(PKNCAconc) != group_vars(PKNCAresults)

Originally posted by @Gero1999 in #451 (comment)