Linear regression example

Data pre-processing phase: (1) reponse transformation, (2) standardization of the continous variables and (3) dummyfication of categorical variables.

Full model: all factors and second order interactions

Low VIF model: all variables with 'high' VIF where removed.

Reduced model: only statistical significant variables. This part also have a regression diagnostics

Final model: only statistical significant variables and no influent ouliers