It is important to ________ ["center", "square", "cube"] the x values before squaring them when we are fitting a quadratic regression model because doing so will _______["reduce", "increase"] the effects of _______ ["multicollinearity", "outliers", "extrapolation"]