Finite-sample distributions of the OLS estimator and t-statistics are complicated,unless
A)the regressors are all normally distributed.
B)the regression errors are homoskedastic and normally distributed,conditional on X1,...Xn.
C)the Gauss-Markov Theorem applies.
D)the regressor is also endogenous.

Question

Quizplus · Accepted Answer

The finite-sample distributions of the OLS estimator and t-statistics are complicated unless the regression errors are homoskedastic and normally distributed, conditional on X1,...Xn. This is because when the errors are homoskedastic and normally distributed, the OLS estimator and t-statistics have a normal distribution with known mean and variance, making hypothesis testing and interval estimation reliable. The other options (A, C, D) are not sufficient conditions for simple finite-sample distributions of the OLS estimator and t-statistics.

Finite-Sample Distributions of the OLS Estimator and T-Statistics Are Complicated,unless