Purpose: The purpose of this paper is to help understand a social phenomenon with the statistical tools we have used. This means choosing a relevant dependent variable and at least two relevant independent variables. This is a methodological paper, which means you’re not required to engage in theoretical debates as you would in a substantive class. Even so, it is always more interesting if the topic and variables you choose are revealing to you and the reader in some way.

Format: 4-5 double-spaced pages with 6 separately labelled sections:
1) Abstract: under 100 words
2) Introduction
3) Theory: Not a traditional theory section, but a short write-up that indicates how you conceptualise the question or puzzle. Tell us what is not known, why it’s important, and what one or more competing explanations may be.
4) Method: This time, don’t include all the steps of the analysis in your write-up. Instead, the “Method” section should give a brief overview of the tests you will perform and indicate that the data have met the relevant requirements/assumptions.
5) Results and Discussion: Present all statistical tests in tables. Ask me or look at published papers for ideas on how to present them. Pasting Stata output is a UUUGE no-no here! Reference relevant results from the tables in the text using precise numbers. Any time you reference a statistic that has a p-value with it, include the p-value in parentheses. Any time you reference a significant relationship in regression, include the (ideally standardised) coefficient and p values. For example,
a. A Gamma test indicated a strong positive relationship between x and y (Gamma = 0.8). A chi-square test further indicated this was a significant relationship (p