2014-15 multiple choice Flashcards
Is the OLS slope coefficient equal to the ratio of sample cov(X,Y) to var(X)
Yes
Is the OLS intercept the value consistent with Y=0?
No
For the OLS estimator to be unbiased, do all gauss-Markov assumptions need to hold?
No
Will the R^2 for a regression always increase when extra variables are added?
Yes
Will the omission of a relevant variable normally lead to bias in the remaining coefficient estimates?
Yes
Does serial correlation in the residual imply bias in the coefficient estimates?
No
Does Heteroscedasticity in the residuals imply that the OLS estimator will be inefficient?
Yes
If the Durbin-Watson statistic is less than two then does this indicate the presence of negative autocorrelation
No
In a bivariate regression is the F statistic equal to the t statistic for the slope coefficient?
No
Does the covariance of two variables X and Y lie in the range -Infinity to infinity?
Yes
In bivariate regression model, The slope coefficient can be calculated as:
∑(Xi-Xbar)(Yi-Ybar)/∑(Xi-Xbar)^2
Is an efficient estimator one for which the variance is the lowest possible in the range of estimators considered?
Yes
If E(error)=0 and X variable is non-stochastic, Does this prove the unbiasedness of the OLS estimator?
Yes
Under Gauss-Markov theorem conditions, does the OLS estimator have the lowest variance in the class of linear unbiased estimators?
Yes
What can you use as a test for heteroskedasticity?
The white test
Can we use the Durbin H statistic when we wish to test for serial correlation and the equation includes a lagged
endogenous variable?
Yes
Is serial correlation when equation errors are related to their own past values?
Yes
Can we eliminate heteroskedasticity in equation residuals by estimating the equation in log-linear form?
Yes
What type of test statistic should we use to test if the slope is equal to one?
B-1/s.e.b
What’s the F ratio?
((Rsq/(1-Rsq))x(N-2)
What should you do before deciding if X has an effect on Y?
Re-estimate the model in difference form
What will give an approximate lower bound for the intercept (a)
a-1.96xs.e.a