Econometrics 2400-PP3EKOa
List of topics:
1. Introduction
a. Subject of econometrics
b. The idea of econometric model
2. Ordinary Least Squares (OLS)
a. Derivation of OLS estimator
b. Properties of regression hyperplane, decomposition of the sum of squares, measures of fit and their properties
3. Interpretation of model parameters
a. Dummy variables
b. Models linear with respect to transformed variables (logarithmic, translogaritmic, spline model)
c. Partial/marginal effects
4. Classical Linear Regression Model (CLRM)
a. Assumptions of Classical Linear Regression Model (CLRM).
b. Properties of OLS estimator in CLRM: expected value and variance.
c. Estimator of linear function of parameters and its variance
d. Making predictions with OLS: prediction variance and variance of prediction error.
e. Efficiency of OLS estimator in CLRM: Gauss-Markov theorem
5. Statistical inference in CLRM
a. Assumptions on the error term: Classical Normal Linear Regression Model (CNLRM)
b. Distribution of OLS estimator in CLRM.
c. Testing the simple and joint hypotheses: test t and F.
6. Diagnostic tests
a. Diagnostic checking. Testing assumptions of CLRM.
b. Tests of:
i. functional form (RESET)
ii. Normality of error term (Jarque-Berra)
iii. stability of parameters (Chow)
iv. homoskedasticity (Breusch-Pagan, White)
v. autocorrelation (Durbin-Watson, Breusch-Godfrey)
7. Fundamental problems of estimation with OLS
a. Omitted variables (intervening variables): empirical example
b. Incorrectly included variables
c. Outliers and erroneous observations
d. Multicollinearity
e. Asymptotic properties of OLS and simultaneity
8. Heteroscedasticity and autocorrelation
a. Causes of heteroscedasticity and autocorrelation
b. Consequences of heteroscedasticity and autocorrelation
c. Generalised Least Squares (GLS)
d. Transformation of GLS estimator to OLS estimator
e. Feasible GLS (Weighted OLS)
f. Robust estimators of variance matrix.
Type of course
Course coordinators
Learning outcomes
The course main objective is to teach students the basic methods used in empirical research in economics. The lecture is to make student familiar with OLS estimator, statistical inference in OLS, diagnostic tests, autocorrelation and heteroscedasticity, simultaneity/endogeneity problem, omitted variable problem, identification of parameters, GLS estimator. The problem sessions are intended to teach students the practical aspects of the applications of the econometric tools mentioned above.
Upon completion of the course student should be able estimate himself linear economic model and to interpret the interrelations between analyzed variables with estimated coefficients of the model. Student should also be able to identify the variables whose influence on other variables is statistically significant and to test the validity of statistical and functional assumption upon which the model is based.
KW01, KU01
Assessment criteria
Final grade is a weighted average of the grades from written exam and problem sessions with weights 2/3 and 1/3 respectively. Students who failed the problem sessions are not permitted to take the exam.
Written exam takes 90 min and consists of 4 theoretical questions, 2 modified exercises similar to the problems in the problem set, and 1 exercise not included in problem set. Theoretical questions are modified versions of the questions given at the end of each lecture. In order to pass the exam student has to solve at least one exercise and answer 2 theoretical questions.
Grading of problem sessions is based in 40% on final test, in 20% on quizzes and activity and in 40% on the grade from the empirical model. The autors of the best model submitted will be exempted from writing the exam but under condition that they have at least 4 from the final test.
Problem sessions are obligatory.
Additional information
Additional information (registration calendar, class conductors, localization and schedules of classes), might be available in the USOSweb system: