Application Project

you will complete a project to demonstrate your understanding of the statistical methods covered in this course, as well as to demonstrate your ability to make such calculations. You will do so by selecting variables of interest from a publicly available dataset, providing an overview of the data therein, creating a hypothesis and null hypothesis, and then testing those hypotheses. You can utilize the panel structure of the data (the same individuals are contacted multiple times during presidential election seasons and their responses are linked to the same respondent), or you can look at variables within a single instance in the panel. I strongly encourage you to use R/R-Studio or SPSS to analyze date for this project (or STATA if you have access already), as these are the most common tools for statistical analysis in the social sciences

What You’ll Need:

You will need to go to electionstudies.org and register for a free account. Then go to electionstudies.org/data-center to select a dataset for analysis. Please note that the code books will give you the question content, and will explain the meaning of the data used in the questions you select. (This won’t always be obvious–perhaps a yes/no (binary) question was asked, but the dataset is likely to contain 0’s and 1’s only, so knowing that a 1 indicates a “yes” is key.)

It is recommended that you also use a statistical analysis software. You can download R (r-project.org) and R studio (rstudio.com) for free Online. Google Sheets or Excel are also occasionally used by statisticians in the social sciences, but these tools are limiting based on processing power and capabilities and increase the likelihood of a mistake in your calculations. you will find a tremendous number of guides and forums Online to assist you in working through the software.

Project Requirements:

You should submit your findings as follows:

The entire project should be submitted in a Word document or a .pdf

Submissions MUST have a works cited to cite your data source, as well as any other resources you consulted (such as the textbooks, lectures, etc.). Failure to include a works cited page will result in an automatic zero

You should format your paper as follows:
Section 1:
Provide a brief (1-3 sentence) description of your hypothesis and the reasoning behind it.
Ex: “Based on the work of Miller (1963) I expect to see that individuals with higher incomes identify as increasingly liberal.
Formally state your H and H0.
Ex: H1: As an respondent’s income increases, that respondent becomes increasingly likely to identify as liberal.
H0: There is no relationship between a respondent’s income and that respondent’s ideology.
Section 2:
In a table, provide the following:
Sample size, averages of variables of interest (be sure to use the appropriate measure for the class of variable you examine), he range of your variables of interest, the variance and standard deviations of variables of interest, as well as any other descriptive factors you think should be included.
Be sure to provide a title for your table, and in a footer in the table, be sure to cite your data source.
Describe all aspects of the table in prose. Point out any irregularities or unexpected findings. This may be a few sentences or a few paragraphs–just be sure to hit on each part.
Section 3:
Explain the test you will be using to test your hypotheses. Define the test and briefly (in 1-2 sentences) explain how the test works. Then justify this test as appropriate for your data/hypothesis. Be sure to justify any decisions you made related to error or significance cutoffs.
In a new table (with a title and a footer with citation information AND an explanation of your error cutoffs), provide the results of your hypothesis test.
Include a figure following the table (also titled, also containing the citation footer with error explanation) that shows your curve and critical regions. (This figure can be hand-drawn, and does not have to be to scale, but should be conceptually and numerically accurate.)
Provide a written explanation of the outcomes of your test. Do you reject the H0? What does this mean in practical terms?
If you reject the null, be sure to add a final explanation of the magnitude of the relationship; if you do not reject the null, explain how you would have measured the magnitude of the relationship.
Section 4
Works Cited
.