Which of the following limitations of the linear probability is not problematic in the event that the only treatment variable in your regression (X or right-hand side), is a binary variable?
A) The errors of the model have to be heteroscedastic.
B) The errors of the model have to be homoscedastic.
C) The predictions might not fall between zero and 1.
D) The coefficients might be negative.

Question

Quizplus · Accepted Answer

The predictions might not fall between zero and 1. This is not problematic in the event that the only treatment variable in your regression (X or right-hand side), is a binary variable because the predicted probability will always be between 0 and 1.

Which of the Following Limitations of the Linear Probability Is