Question

1. When the error-terms are heteroskedastic, then:

a) WLS is more efficient than OLS in large samples, if the functional form of the heteroskedasticity is known.

b) OLS coefficients are biased.

c) OLS is still BLUE but t and F distributions are invalid.

d) there are only two solutions: either use WLS if the functional form of the heteroskedasticity is known or use GLS if the functional form of the heteroskedasticity is known.

e) None of the above

2. Which of the following is TRUE?

log(wage) = β0 +β1female +β2educ +β3female*educ +β4exper +β5exper*female +β6exper2 +u

a) Rejecting H0: β1 = β3 = β5 = 0 is evidence against existence of wage gap between male and female.

b) Failing to reject H0: β1 = β3 = β5 = 0 is evidence for existence of wage gap between male and female.

c) Rejecting H0: β1 = 0 and failing to reject H0: β3 = β5 = 0 is evidence against the eect of education and experience in reducing wage gap between male and female.

d) Rejecting H0: β1 = β3 = β5 = 0 is evidence for existence of wage gap between male and female.

e) Failing to reject H0: β1 = 0 and rejecting H0: β3 = β5 = 0 is evidence for non-existence of wage gap.

Answer 1

1. When the error-terms are heteroskedastic, then:

d) there are only two solutions: either use WLS if the functional form of the heteroskedasticity is known or use GLS if the functional form of the heteroskedasticity is UNknown.

2. Which of the following is TRUE

log(wage) = β0 +β1female +β2educ +β3female*educ +β4exper +β5exper*female +β6exper2 +u

d) Rejecting H0: β1 = β3 = β5 = 0 is evidence for existence of wage gap between male and female.

Please revert back in case of any doubt.

Please upvote. Thanks in advance.