Econometrics

Subject

Multiple Linear Regression.

(i) The models with two and p explanatory variables (using matrix representation): the detection of a linear relation using a three dimensional scatter diagram (for p independent variables see iii below), the model specification using theoretical information and empirical evidence for the choice of the explanatory variables, interpretation of the partial regression coefficients, point estimation of the parameters using the method of Ordinary Least Squares (OLS), standard errors of the estimators, partial elasticities of the dependent variables with respect to the independent variables and their interpretations, the classical assumptions for the “best” estimators using OLS, interval estimation of the parameters, hypotheses testing for the significance of each regression coefficient, overall significance test with the Analysis of Variance, the coefficient of multiple determination and the adjusted coefficient of multiple determination, Mallow’s C_p the Extra Sum of Squares Principle, partial correlations, indicator variables and predictions.

(ii) Model Selection Methods: Backward Elimination, Forward Selection, Stepwise Regression and All Possible Regressions.

(iii) Violations of the Classical Assumptions and Remedies: model specification errors (omitted and irrelevant variables, incorrect functional form etc.), heteroscedasticity, collinearity (or multicollinearity), autocorrelation, lack of normality of the error term, use of random regressors. The possible causes and consequences of each violation, and methods for detection; residual analysis, graphical techniques, statistical tests, Variance Inflation Factors and Condition Indices (for the magnitudes of eigenvalues of the X^TX matrix). Various remedies including data transformations, and the methods of Generalized Least Squares and Weighted Least Squares for the estimation of parameters, in the regression models with AR(1) and heteroscedastic errors respectively. 

Educational Aims

Student Evaluation

Written exams at the end of the semester.

Bibliography

1. Draper, N. και Smith, H., Εφαρμοσμένη Ανάλυση Παλινδρόμησης, Παπαζήση, Αθήνα 1997 (in Greek).
2. Johnston, J. και DiNardo, J., Οικονομετρικές Μέθοδοι, Κλειδάριθμος, Αθήνα 2004 (in Greek).
3. Κάτος, Α., Οικονομετρία (Θεωρία και Εφαρμογές), Ζυγός, Θεσσαλονίκη 2004 (in Greek).
4. Χρήστου, Γ., Εισαγωγή στην Οικονομετρία, Τόμος Α, Gutenberg, Αθήνα 2004 (in Greek).