Is it only me, or are we getting more and more questions about normality of predictors? Here is another one. Do we have a canonical thread on whether predictors need to be normally distributed?
@StephanKolassa where are they getting the fact that we need normality for predictors?
In any case, you seemed to be right that, of late, there has been uptick of such posts. And I can think of many good posts that could be used as reference/canon though I have to check for a definite one.
