Econometrics and Regression Results Assignment | Online Assignment

Project Handout 3

Some Overall Comments and Advice

Outline:

• Literature Review
• Methodology
• Regression Results (style and content)
• Final Comments

Literature Review

• Do NOT write more than a paragraph about each of your articles
• Answer four questions in this paragraph:

What is the topic being studied?

How is the analysis being conducted by the author(s)?

What results were obtained?

How does this article relate to YOUR project?

• Please place full references in a bibliography, not in a footnote.
• Finally, a style suggestion: when talking about an article, do not bother to write out the entire name of the author and article. Just put the author’s last name, and the year of the publication (in parentheses).  For example, instead of ‘Bob Friedman, in his book ‘The Efficiency of Civil War Tax Collectors in rural southeastern North Carolina’, analyzes…’ try writing ‘Friedman (1993) analyzes possible reasons for…’

Methodology

I would like several items to be included in your methodology:

• Data sources (where you obtained your data)
• Definitions of your variables (this can be put into a table, if you’d like)
• Econometric model to be estimated (the big one, not all the smaller technical regressions). If you want to get fancy with your notation, try using Equation Editor, which is located in INSERT>>OBJECT. MathType is also a useful tool.
• Expected signs of coefficients
• Tests to be conducted (not all may apply to your paper). These will fall into three categories:

Individual Coefficient Tests (to check statistical significance, economic theory)

Joint Hypothesis Tests (overall significance of regressions, seasonal effects, etc.)

Tests for multicollinearity, Serial Correlation or Heteroskedasticity

Regression Results

This section seems pretty obvious at first, but questions may emerge as to what to include.  Here are some rough suggestions.

l Include at least some of your Stata printouts.

l Do NOT include EVERY SINGLE regression output that you ran! Include your basic regression model output.  Include a few others (no more than two or three), if you are struggling with whether or not to include several variables.  Also include any regressions you had to run to correct for heteroskedasticity or serial correlation.  (Just a quick explanation in your paper, along with the F-stats, critical values, etc. will suffice for these).  So, that comes out to four or five estimation outputs, tops.

l Finally, be sure to explain your printouts.  Don’t go overboard here (that is, don’t comment on ‘Mean Dependent Variable’), but at the very least, talk about the statistical significance of each slope coefficient and the overall significance of the regression!  If done well, this explanation will not take up much space.  Remember, this paper is to be only ten to fifteen pages long.

Final Comments

A few comments about turning in the completed project:

When turning in your project, you must also submit the following:

An electronic version (e.g. MS Word or PDF) of your project paper.

Final data set used (Stata or Excel format; screenshot of first page if data size>20MB).