ABSTRACT
All concerns and simple remedies applicable to simple regression models apply here: Transformations of the response or explanatory variables to improve fit; transformations of the response variable to address violations of the normality assumption; and transformations of the response variable to stabilize variances. For the most part, the methods and functions described for simple regression models work just fine here, but the presence of two or more explanatory variables can make identifying a suitable transformation of the explanatory variables tricky.
