Under the least squares assumptions for the multiple regression problemâ (zero conditional mean for the errorâ term, all LaTeX: X_iXi and LaTeX: Y_iYi being i.i.d., all LaTeX: X_iXi and LaTeX: \mu_iμi having finite fourthâ moments, no perfectâ multicollinearity), the OLS estimators for the slopes andâ intercept:
A. have an exact normal distribution for nâ > 25.
b. are unbiased and consistent.
c. have a normal distribution in small samples as long as the errors are homoskedastic.
d. all BLUE