Suppose the true regression model is: Y =B + BX + BX3j + Uj 1 of...

 

Transcribed Image Text:

Suppose the true regression model is: Y₁ =B₁ + B₂X₂₁ + B₂X3j + Uj 1 of 3 Now, consider the following advice: "When the explanatory variables X₂ and X, are correlated, the variance of b, is larger than it would be if X, and X, were uncorrelated. Thus, if you are interested in P₂, it is best to leave X, out of the regression if it is correlated with X₂." What do you think of this advice? Suppose the true regression model is: Y₁ =B₁ + B₂X₂₁ + B₂X3j + Uj 1 of 3 Now, consider the following advice: "When the explanatory variables X₂ and X, are correlated, the variance of b, is larger than it would be if X, and X, were uncorrelated. Thus, if you are interested in P₂, it is best to leave X, out of the regression if it is correlated with X₂." What do you think of this advice?

Answer rating: 100% (QA)

Here are more indepth workings on why removing a correlated explanatory variable can help estimate o... View the full answer