Consider the following regression model. log() = Bo + Bi X log(xi) + B2 x log(x2) + € Now consider two conditions: Condi
Posted: Wed Dec 15, 2021 10:24 am
Consider the following regression model. log() = Bo + Bi X log(xi) + B2 x log(x2) + € Now consider two conditions: Condition A: log(xi) + log(x2) = 1 for every observation, Condition B: x1 + x2 = 1 for every observation. O Condition B leads to perfect multicollinearity but not condition A. Both Condition A and Condition B lead to perfect multicollinearity. Condition A leads to perfect multicollinearity but not condition B. Neither Condition A nor Condition B lead to perfect multicollinearity.