Guidelines for empirical paper

These guidelines describe the format you should use, and content you should include, when
writing your paper. Deviations from them will negatively affect your paper grade, but
following them will positively affect your paper grade (!).
1 Introduction
1. Statement of research question (What is relationship you wish to examine? What is
your main independent variable and what is your dependent variable?) Note: this
must appear in 1st or 2nd paragraph
2. Motivation (Why is your question important? Why do we need to understand the
relationship between your main independent variable (X) on your dependent variable
(Y )?)
2 Literature Review
All empirical papers include some sort of literature review. In this section of the paper
you are basically laying out what is currently known about your topic (or topics similar to
yours). For this assignment I want you to:
1. Cite at least 3 economics papers
2. For each paper, write 1-2 paragraphs describing what question the paper seeks to
answer, what data is used, and what the key findings are. Discuss also how these
results relate to the question you are attempting to answer in your paper.
**** The BELOW should not be included in your literature review, but I want you to think
about these things as you read the papers in your literature review. ****
This is also an opportunity for you to be able to learn a little more about how others
think about your topic. In particular, what other variables have they included in their re-
gressions? That is, what other factors have other researchers thought should be controlled
for?
**** The ABOVE should not be included in your literature review, but I want you to think
about these things as you read the papers in your literature review. ****
1
3 Data and Descriptive Statistics
All empirical papers also provide a section in which they describe the data they will use in
the analysis.
The backbone of this section will be your Table 1 (see STATA table1 example.do file for
instructions on how to make this). In this table you will present basic descriptive statistics
of the key variables in your data.
Two of these variables must be your X and Y variables.
What other variables to include? There are two ways to think of additional variables to
include in your Table 1: (1) variables that help describe the basic profile of the observations
in your data (e.g. if an observation in your data is an individual, you might provide
information on age, gender, education, income, ethnicity, etc); and (2), additional variables
you will include as controls in your regressions.
For this assignment I want you to:
1. Describe the data you will use to test your research question, including but in no
particular order:
(a) Describe the source of the data
(b) Describe which variables it contains
(c) Describe the timeframe for the data
(d) Describe the unit of observation (an individual, country, year? Etc)
(e) If you have made choices to limit your analysis to certain years, companies,
countries etc, justify your choices (you may use the theory to justify these choices
or the availability of the data etc)
2. Cite and refer to your descriptive statistics table (Table 1) within this section to help
your discussion of describing the data
4 Empirical Analysis
1. First run a simple regression of your dependent variable (Y ) on your main independent
variable (X)
(a) Describe the regression results – interpret the coefficient estimate for X, explain
whether it is statistically significant and at what significance level
2. Next move to the multiple regression framework
(a) Think about Omitted Variable Bias – does the simple regression you ran in (a)
give an unbiased estimate of the relationship between X and Y ? Think of some
(at least one) variable that, if omitted, potentially causes your estimate in (a)
to be biased. Go through the logic of what direction the bias should be, and
provide a discussion. Then include that variable in the regression.
2
(b) Think about what other variables you might want to add as controls and what
other modifications to the regression equation you might want: i.e. which ad-
ditional variables to include, interactions, and functional form. Carry out those
changes and report your findings.
3. For each additional regression:
(a) Interpret your findings in words; do your results support your main hypothesis?
(b) Explain why the coefficient of interest (the coefficient on your main X variable)
does or does not change between regressions
(c) Interpret the coefficients on the additional X variables – i.e. how do these vari-
ables relate to the dependent variable Y ? Also, particularly in the case of in-
teraction terms (if used), how do these variables alter the interpretation of the
relationship between your main X variable and Y ?
(d) Discuss which regression equation fits the data best
4. Finally, discuss shortcomings of your research (For example, are there further omitted
variables that you don’t have any data for? If so, how would this affect the estimated
coefficient of interest? Is it biased? In what direction – larger or smaller? Why? )
5. Make a table to show all of your regression results (see STATA main table example.do
file for instructions on how to make this).
5 Conclusion
Summarize your topic and findings
6 Works Cited
Format: follow format used in the bibliography of one of your literature review papers.
7 Tables
1. Descriptive statistics table
2. Main regression results table