es: The first step would be evaluation of correlation between predictors (deletion of 1 of those predictors who are significantly correlated)
viņš: This not correct!!!! Who told you that you must exclude by a model one predictor due the fact that is highly correlated with another? It is not correct at all!