Econometrics is a tool for proving hypotheses deduced from economic theories. Estimation can be done very simply by least squares method, and hypotheses can thus be tested. However, this simplicity often hides presumptions which guarantee credibility of the test results. Posterior evaluation to examine effectiveness of the presumptions is demanded. If some presumptions are found to be failed, more sophisticated estimation methods and testing methods should be applied.
Econometrics is also a tool for forecasting. However, in the forecasting process, the fact that estimated parameters and predicted values are stochastic variables is often set aside. We should learn that variance of the forecast is equally important to its value.
Starting from the Classical Simple Regression model, we study multiple regression model with matrix algebra, and methods to test hypothesis either simple or combined for examining structural changes. Further we study how to deal with problems caused by model specification, i.e. selection of explanatory variables and functional form, and those by multiple collinearity.
Regression models are then extended as generalized classical regression model to treat with problems caused by heteroskedasticity and serial correlation.
Finally Neoclassical regression models are studied to handle stochastic explanatory variables, with both instrumental variable method and generalized moment method being employed.
Least Sqaures Method, Regression Model, Classical Regression Model, Matrix Algebra, Hypothesis Testing,Structural Change, Model Specification, Multiple-collinearity, Generalized Classical Regression Model, Heteroskedasticity, Serial Correlation, Neoclassical Regression Model, Stochastic Explanatory Variable, Generalized Moment Method.
Intercultural skills | Communication skills | Specialist skills | Critical thinking skills | Practical and/or problem-solving skills |
---|---|---|---|---|
- | - | ✔ | - | - |
At the beginning, solutions to some exercises related to the preceding class are explained. Scheduled contents are then lectured on. Exercises of the related chapter are given as homework.
Course schedule | Required learning | |
---|---|---|
Class 1 | Conditional Expected Value and Line Fitting | End-of-chapter exercise related to scheduled contents of the class |
Class 2 | Classical Bivariate Regression Model | |
Class 3 | Test of Parameters and Forecast | |
Class 4 | Multivariate Regression Model | |
Class 5 | Matrix Algebra for Linear Regression | |
Class 6 | Classical Multivariate Regression Model | |
Class 7 | Hypothesis Test | |
Class 8 | Application of Multivariate Regression Model | |
Class 9 | Test of Multiple Hypotheses: Constrained Regression and Test of Structural Change | |
Class 10 | Model Specification | |
Class 11 | Multi-Collinearity | |
Class 12 | Generalized Classical Regression Model: Heteroskedasticity and Serial Correlation | |
Class 13 | Generalized Classical Regression Model: Estimation of Simultaneous Equations | |
Class 14 | Neoclassical Regression Model: Stochastic Explanatory Variable and Probability Limit | |
Class 15 | Neoclassical Regression Model: Instrumental Variable and Generalized Moment Method |
Asano & Nakamura(2009) "Econometrics" Second Edition, Yuuhikaku (in Japanese)
No particular references.
40％ by Short tests and homeworks, and 60% by Final Examination.
Prerequisite
IEE.A204 ： Probability for Industrial Engineering and Economics
IEE.A205 ： Statistics for Industrial Engineering and Economics