You really should consider eliminating as many variables as possible. The idea is to obtain good data and then incorporate all the variables, determined individually, when interpreting the data. If you are testing for the film's characteristics, you need to separate it from the optical system, and that means contacting.
very true, but i find contacting difficult to do. that said i once made my self an adaptor to fit an lftaking lensto my enlarger,which alloed me to use it's shutterand expose filmcontacts from 1/250s to 1s .it worked wellwhen i used it to make msks for unsharpening.