Introductory Analysis of Political Data
Professor Mikel Wyckoff
Zulauf 403 753-7056
Office Hours: MWF 11-12:00 & by appointment
This course provides an introduction to basic statistical methods frequently encountered in the social sciences. Topics we will cover include: (1) descriptive statistics (for example, means and standard deviations); (2) statistical inference (random sampling and hypothesis testing); and (3) measures of relationship and association. Students will also get an introduction to SPSS for Windows, a widely used data analysis program.
Although I personally enjoy statistical methods very much, as a realist I suspect that at least some of you approach this class with something less than enthusiasm. Let me assure everyone that the course is not that difficult (almost everyone who is willing to put in the time and effort can expect to earn a respectable grade). Let me suggest, furthermore, that if you give the material a fair chance it is likely that you will find these research tools useful (at the very least) and even a tiny bit interesting. For those experiencing genuine fear and anxiety, the textbook's section on “Overcoming Math Anxiety” may be helpful.
The following books are required and should be available at local bookstores. If not, they are readily available from online sources.
· John H. Kranzler, Statistics for the Terrified, Prentice Hall, 4th ed., 2007.
Michael S. Lewis-Beck, Applied
Regression: An Introduction,
In this course we will compute a few problems by hand but we will primarily make use a statistical software package called SPSS when we analyze data. Over the years, SPSS has become increasingly user friendly and I believe that all of you can learn the basics of the program simply by watching my demonstrations in class. Those who wish to become even more proficient at using SPSS, however, may wish to consult one of the following guides:
· Philip H. Pollock III, An SPSS Companion to Political Analysis, CQ Press, 2008.
· Marija Norusis, SPSS 16.0 Guide to Data Analysis, 2/E, Prentice Hall, 2009.
Exams. There will be three examinations--two midterms and a final. Together the three exams will constitute 70% of your final grade, with Exams 1 and 2 worth 20% each and cumulative Exam 3 worth 30%. A small component of each exam (typically dealing with definitional issues) will be closed book, closed notes, but when working data analysis problems you are welcome to consult your textbook and your notes if need be. An inexpensive electronic calculator will probably be useful when taking the exams.
Computer Exercises. The remaining 30% of your grade will be determined by your performance on computer assignments which will require the use of SPSS statistical software available in the main SOCQRL (in DuSable Hall) and in the SOCQRL satellite lab that we have on the first floor of Zulauf (course fees enable you to use these well-furnished labs). No one should feel intimidated by these assignments. No prior knowledge of statistics is required, the software is user friendly, and experienced lab assistants will be available (especially in the main SOCQRL) to help with the assignments when needed. Of course, all students must do their own assignments. This means learning to use the required software on your own, conducting your own analyses, obtaining your own output, and writing your own reports.
Take-Home Problems. Throughout the semester I will assign sets of problems for you to work outside of class. I expect you to do these problems, and we will spend a good deal of time on them in class, but I will not collect them and grade them. Similar computational problems will appear on the midterm and final exams, so devoting significant time to the take-home problems is well justified despite the fact that they will not be graded.
COURSE OUTLINE AND READING ASSIGNMENTS
I. PART ONE: Descriptive Statistics
· Introduction and basic math review – Kranzler, Appendix A and Ch. 1-2.
of Measurement, Frequency Distributions –
Measures of Central
Tendency and Variation –
Standardized Scores –
EXAM I: Monday, October 5
II. PART TWO: Data Analysis – Using Independent Variables To Explain Dependent Variables
A. Analyzing Categorical Data with Crosstabs
1. Introduction to crosstabs; relationships and measures of association
· The Crosstabs Procedure. – See handout, “How to Read a Crosstab” that I placed in your mailbox.
· Measures of Association for Nominal and Ordinal Data. – See untitled handout that I placed in
your mailbox. For now read pp. 144-151, starting with “Measures of Association.”
2. Hypothesis testing (significance testing) with crosstabs; holding “third” variables constant
· Random Sampling, Inferential Statistics and the Concept of a Sampling Distribution
Kranzler, Ch. 10.
· Hypothesis Testing I: The chi-square test – Kranzler, Ch. 13.
· Using Control Variables with Crosstabs – Reading TBA.
B. Analyzing Interval Level Dependent Variables and Categorical Independent Variables
1. The t-test – Analyzing differences of means observed in two groups
· Additional Thoughts on Sampling Distributions; Standard Errors.
· Hypothesis Testing II: The t-test – Kranzler, Ch. 11.
2. ANOVA (Analysis of Variance) – Analyzing differences of means observed in three or more groups
· Hypothesis Testing III: The F-test and Analysis of Variance – Kranzler, Ch. 12.
EXAM 2: Monday, November 16
C. Analyzing Two or More Continuous Variables – Bivariate and Multivariate Regression Analysis
· Measures for interval data: Pearson’s r and the slope of the regression line
Kranzler, Ch. 8. Lewis-Beck, pp. 9-25, 43-47.
· Significance tests and confidence intervals – Lewis-Beck, pp. 30-37.
· Multivariate analysis – Lewis-Beck, pp. 47-58.
· Dummy variable analysis – Lewis-Beck, pp. 66-71.
EXAM 3: Monday, December 7,