Descriptive Statistics and Comparing Variables and Controlled Analysis (Example Here)
Assignment Goals - The goal of this paper is to demonstrate that the student understands the processes involved in social science research, including hypothesis formulation, formulating a literature review, reporting statistical output, doing a controlled analysis, and deriving conclusions from the statistical reports. This assignment gauges the student’s ability to apply the research methods and procedures learned in class.
Use Pollock Chapters 2, 4, and 5 as a reference guide to complete this assignment. A paper at this level should be at least 12 pages in length.
Write a research paper about two variables in which you are interested. (You should continue to use the variables from paper #2.) The paper will consist of an introduction, a short literature review (cite at least FOUR academic articles from an academic database, including journals and academic books), hypothesis to be tested, a short discussion of the variables’ summary statistics (central tendency and dispersion), statistical procedures you will run to compare the variables, tables and appropriate graphs of the results, a discussion and implication of the results. You should also have a controlled analysis as part of this paper. The control variable should be discussed in your literature review.
Reference the following link for a discussion on how to successfully structure your paper:
Weingast, Barry R. 1995. "Structuring Your Papers."
Paper format should conform to the American Political Science Association’s “Style Guidelines”.
Also, consider the Ten Commandments of PS Writing.
Format and description
Answers the question: Why am I studying these variables? Why is this important? What do you expect to find? State why you think this relationship is important. Is the relationship positive, negative, a correlation, or a spurious relationship? Why is it one and not the others?
In this section, it can be appropriate to use popular media citations (like newspapers and magazines) and citations from government sources (the CDC, the FBI's UCR, etc.) to provide context for the relationship or question you are asking. It is always appropriate to use academic citations from JSTOR or Google Scholar (http://scholar.google.com/ (Links to an external site.)).
Answers the question: how have others studied this topic in the past?
Using JSTOR find two related research articles and cite the appropriate in your paper. Eg, what type of research design was used? What was the data? What were the authors' findings/conclusions? How does it relate to the variables you are studying? What other factors can influence the relationship? What is your control variable? What does the literature say about the control variable and how it might affect the dependent variable? What do you expect to find?
Be sure and write in the appropriate style. Use the guidelines set forth by the APSA writing manual and the handout found HERE and HERE.
This section should end with an explanation linking past research the hypothesis you proposed. Make sure to use the template Pollock offers: in a comparison of (unit of analysis)...
Data and Methods:
Answers the question: What data set am I using? How many people answered the relevant questions? What are the variables and how did they ask/measure the phenomena? What are the measures of central tendency/dispersion?
If you report SPSS output DO NOT use tables from SPSS. Make your own tables in your Word processor and only report the relevant numbers. DO use relevant bar charts and histograms when appropriate.
Answers the question: what statistical procedure are you using? Crosstabs/comparison of means? Why? Explain the tables and graphs you produce using SPSS. Does it prove or disprove your hypothesis?
Do not use the tables created by SPSS. Create your own in your word processor. You can use line charts and bar charts imported from SPSS, though. Report the results of your controlled analysis.
Answers the question: What are the implications of the findings? Why should we care? Discuss what would make your research better? Are there other variables that may change the nature of the relationship? Explain how that may be the case.