You found a serious multicollinearity problem in your regression model. It turns out that x1 and x2 are highly correlated to each other. Therefore, you need to drop one of these variables. You also found that the correlation of y and x1 is 0.81 and the correlation of y and x2 is 0.75.

Which of the variables will you drop?

a. y

b. x2

c. none of them. dropping a variable is not a remedy for multicollinearity.

d. x1

Question

Business

Shelby Riley

Today, 08:18

You found a serious multicollinearity problem in your regression model. It turns out that x1 and x2 are highly correlated to each other. Therefore, you need to drop one of these variables. You also found that the correlation of y and x1 is 0.81 and the correlation of y and x2 is 0.75.

Which of the variables will you drop?

a. y

b. x2

c. none of them. dropping a variable is not a remedy for multicollinearity.

d. x1

+1

Answers (1)

Know the Answer?

Brooklyn Mata · Answer 1

Answer: b. x2

Explanation:

Now, Multicollinearity results from a linear association between 2 independent variables in a multiple regression model. We have perfect multicollinearity if, for example, the correlation between two independent variables is equal to 1 or - 1.

We know the following from the text,

Correlation (y, x1) = 0.81

Correlation (y, x2) = 0.75

To resolve the issue of Multicollinearity, a variable needs to be dropped. Usually it is the variable that has a weaker correlation with the Dependent Variable.

In this case that is x2 which has a lower correlation with Y of 0.75 compared to the 0.81 that x1 has with Y.