2531.18 - Econometrics
Pass grade in Mathematics B at the upper secondary level. Skills and knowledge from or equivalent to those taught in the courses Probability and Statistics, Mathematics 1 and Mathematics 2 is an advantage.
The aim of the course is to provide an introduction to the nature and use of empirical investigation in economics. In addition, the course aims to provide insight into the assumptions underlying these empirical investigations, as well as the problems that may arise in such investigations.
The course examines in detail the multiple linear regression model in the context of cross sectional data. It considers the estimation technique Ordinary Least Squares (OLS), as well as crucial properties like unbiasedness and consistency of OLS. The course also examines situations where the classical assumptions do not hold. It emphasises the problems of heteroscedasticity and correlation between the error term and one or several explanatory variables.
Learning and teaching approaches
New material is presented in lectures and classes. In classes, students solve and present exercises. STATA (or equivalent program) is used extensively.
Having successfully completed the course, the student will have knowledge of the following: - simple linear regression model and the classical assumptions underlying the model - multiple regression model and the assumptions underlying the model - statistical inference in the context of regression analysis - when it is appropriate to talk about a causal relationship between variables. Having successfully completed the course, the student will be able to: - derive simple estimators and statistical properties of these estimators (such as unbiasedness and consistency) - use different estimation techniques, where emphasis is placed on Ordinary Least Squares (OLS), Weighted Least Squares (WLS), Instrumental Variables (IV) and Two Stage Least Squares, and discuss when it is appropriate to use these different estimators. - test various hypotheses concerning the population parameters - calculate and interpret p-values and confidence intervals in the context of regression analysis - choose between different functional forms - use dummy variables - evaluate whether the underlying assumptions are satisfied, and make appropriate changes when they are not satisfied
3 hour written exam. Students may use textbooks, notes and data programs during the exam. They may not use the internet. Students are required to pass a minimum of two assignments in order to attend the exam.
Herit Vivi Bentsdóttir Albinus