ABSTRACT
As we have seen from the different multiple regression analyses we have performed thus far, in some ways it may seem like we are on a fishing expedition trying to discover a useful relationship between the response and predictor variables. Not only do we need to find the appropriate model that describes the relationship between the set of predictor variables and the response variable, but we also need to be sure that the regression assumptions are reasonably met, and that we have not included predictor variables in the regression model that are highly correlated with each other.
