1. Title of the Paper
2. Introduction
a. Why do you think the economic problem or issue you are planning to examine is so important?
b. Review of literature: Review the past work on this or related issues by other researchers/economists. Library is an excellent source for scholarly journals and articles. You must cite at least two journal articles in your paper. You can also search internet.
3. Methodology
a. Theoretical model: discuss the theory or hypothesis related to the economic issue you are going to investigate. Review of literature will be of great help to establish the theoretical model.
b. Econometric model: Specify the econometric model based on theory. Explain all the the variables included in the model clearly and in detail.
c. Data : Clearly specify
i. Sources of data: Internet and libraries are rich sources of data that you might be looking for. To search for the data availability in the library, you may want to contact:
ii. Number of observations: At least 30
iii. Number of independent variables should be at least 2.
4. Estimate the model
Check for the following violations of CLNRM Model:
a. Normality
b. Variable Selection
c. Multicollinearity
d. Heteroscadasticity
e. Autocorrelation
f. Specification bias
5. Result and Discussion
a. Write the estimated model along with all the relevant statistics.
b. Discuss the results in detail
6. Summary
Summarize your findings (results) and explain if they support or not the economic issue you are investigating. Also, write down the limitations of your study.
7. Bibliography: At least two citations should be made from economics scholarly journals.
8. Appendix: Copy of the Data set you used in the study. Also, a copy of the Stata computer output of your estimated regression model and any correction that you made (multicollinearity, heteroscedasticity, autocorrelation etc.)
9. Due Date: A week before the final Exam. Also, send me a copy as an attachment through e-mail.
Features of a good econometric model:
An econometric model should be specified on the basis of economic theory, past experience and personal intuition.
The sample data must be fitted to the model reasonably well, which means that a high R2 value is desirable and majority of the independent variables should be significant.
A single equation model (that we are dealing now) must meet all the assumptions of Classical Linear Regression Model (CLRM):
The model is free from autocorrelation, heteroscedasticity, multicollinearity, and specification bias problems.
The error tem u follows normal distribution.
Coefficients that have been estimated must have the expected signs.
A model must have a strong forecasting power.
There must be a proper functional form between dependent and independent variables. In other words, the relationship between dependent and independent variables must be specified (linear or non-linear) as per actual relationship.