Explain which are the difference between cross section, time series and panel data. If you could choose your data set structure, which one would you prefer?

Question 1-

Under the Gauss-Markov assumptions OLS is said to be `BLUE’.
Which is the difference between a BLUE estimator and an MVUE estimator? Which one would you prefer to use?

(5marks)

Question 2-

Name three estimation methods that deliver consistent results in the context of linear regression models. Which of these estimators requires the strongest assumptions?

Which are the most important properties of an estimator? Discuss and explain why we can work with biased estimators but we cannot work with inconsistent estimators.

(5 marks)

Question 3-

Explain why we need to assume that the Gauss Markov assumptions hold. Discuss whether in applied research questions using data this is the case.

(5marks)

Question 4

Consider the EViews software used during the course and explain how we can change the variance covariance matrix of the errors to account for the presence of heteroscedasticity and for the presence of serial correlation.

What would you do if you had both?

(5marks)

Question 5

Suppose that you ‘d like to estimate the variance of the dependent variable, how would you achieve such goal? Feel free to discuss EViews command if you like.