Define Multicollinearity in the following terms: a. In what type of regression is it likely to occur? b. Why is multicollinearity in a regression a difficulty to be resolved? c. How can multicollinearity be determined in a regression?. d. If multicollinearity is found in a regression, how is it eliminated? Solution Define Multicollinearity in the following terms: a. In what type of regression is it likely to occur? Multicollinearity exists in almost all regressions, but is only a big problem when some of the independent variables are highly correlated. So, if you have many independent variables, there\'s a good chance some of them will be highly correlated. b. Why is multicollinearity in a regression a difficulty to be resolved? Multicollinearity causes the coefficients of the independent variables in question to fluctuate wildly when any of them is put into, or taken out of, the regression. A coefficient tells you the impact of changing one \"X\" variable when all others are HELD CONSTANT. But that is impossible when the indepedent variables are themselves highly correlated: you can\'t change one without changing the others, at least not easily. c. How can multicollinearity be determined in a regression?. One can examine the CORRELATION MATRIX of the independent variables. Although there is no perfect rule, once the correlations go up to .5-.6 or more, you are very likely to have collinearity problems. You can also see whether the coefficient on one X variable changes a lot when another one is removed. d. If multicollinearity is found in a regression, how is it eliminated? The easiest way is to remove all but one of any set of X variables that are highly correlated with one another. For example, you wouldn\'t put the number of bedrooms, number of bathrooms, and square footage of a house as separate independent variables, since they tend to be VERY highly correlated. You\'d try to remove two of them..