I figured out step 1 all by myself, so proud. Totally skipped over step 2 because I'm silly. Collinearity is only mentioned in every single predictive article / training / help reference out there. The Chi-Sq effect had me all tripped up. It's referenced for individual variables in the distribution analysis, but not for overall models.
If only I had taken a step back and realized we were comparing two models where one is the subset of the other, the very use case for a nested test. How often do you get to use a nested test? That depends, how often do you study for the expert exam?
😏