Non-linearity partial regression plots.


Fortran must die
I am hoping someone has worked with partial regression plots before which are supposed to be better to detect non-linearity than bivariate scatterplots.

I ran these for some of the key variables (in theory all but race are interval, but because of policies at the place I work pay for services sometimes has a few distinct levels, that is nearly everyone gets paid 8-21 rates)

I think none of these suggest non-linearity, but I have limited practical experience with them.