Question

state if the statements are true or false and give a reason.

v. In the presence of heteroscedasticity, ordinary lease squares (OLS) is an inefficient estimation technique and this causes t tests and F tests to be invalid.

vi. Heteroscedasticity can be detected with the Chow test.

vii. In the presence of autocorrelation, ordinary lease squares (OLS) produce unbiased estimates and hence t tests and F tests are still valid.

viii. One problem with the use of a lagged dependent variable as an explanatory variable is that it always gives rise to autocorrelation.

Answer 1

v) True

OLS estimator is biased as it will overestimate or underestimate the variance of ₂

so no longer rely on T-test & F-test

vi) False

as heteroscedasticity is deducted by Graphical method , park test , Gleiser test, white's test and goldfeld quandat test

vii) False

in autocorrelation, OLS is not blue, as no minimum variance ( it is unbiased, linear but not efficient) so T-test and F-test are not reliable as the estimated variance is biased

viii) True

as in this case, D-W test cannot be used so that we have to use Durbin’s h-test