Sociology 526B Spring 2001

Quantitative Methods in Sociology



Professor Linda Andes

Email: andes@siu.edu

Office: Faner 3425

Phone: 453-7622

Office Hours: Monday, Wednesday, Thursday; 3:30pm to 5:30pm or by appointment





Objectives: This course will acquaint students with advanced statistical methods used by sociologists and other social scientists. I expect successful students to become both 1) intelligent consumers of statistical information and 2) able producers of regression analyses using SPSS. Furthermore, students should be equipped with the skills necessary to successfully complete the statistics component of the Sociology department's comprehensive exam.



Description: The majority of the course will cover basic topics in multivariate regression: basic linear models, the use of categorical independent variables, non-linearity, model building, and regression diagnostics. We will also study logistic regression: an advanced method for handling categorical dependent variables; and introduce structural equation models.



Pre-requisites: Students must have completed SOC 526a or an equivalent course. This means that I expect you to be generally competent in the areas of descriptive statistics, hypothesis testing including t-tests, chi-squared and one-way analysis of variance, and bivariate regression. This course does NOT require familiarity with calculus.



Required Texts: Allison, Paul D. 1999. Multiple Regression: A Primer. Thousand Oaks, CA: Pine Forge Press.



Pampel, Fred C. 2000. Logistic Regression: A Primer. Thousand Oaks, CA: Sage Publications.



McClendon, McKee J. 1994. Multiple Regression and Causal Analysis. Itasca, IL: Peacock Publishers.







Grading: Homework 35% Due dates will be announced in class

2 Quizzes 15% Monday, March 5th and Monday, April 23rd

Midterm Exam 25% Due on Wednesday, March 21st

Final Exam 25% Due on Wednesday, May 9th









Homework: Learning statistics is not a spectator sport: the only way to learn the material is to do statistics. There will be approximately 7 homework assignments. The majority of the homework will be computer problems using SPSS and writing reports with an interpretation of the results. Homework assignments should be written using a word processor, and double-spaced. Regarding tables, you have three choices: 1) cut-and-paste the relevant tables from your SPSS output into your text; 2) create new tables in your word processing program; or 3) reference the relevant table in your SPSS output using unique table titles or numbers and accurate page numbers. (Examples will be given in class.)



Homework will be graded on a 3 point scale: with a "3" indicating an exceptional level of performance, a "2" indicating the expected level of performance and a "1" indicating an inadequate level of performance. Grades for late assignments will automatically be reduced by one point.



You are encouraged to HELP each other on homework assignments, but I expect the work you hand in to be your own. Of course, if you have questions or need assistance, I also encourage you to see me during my office hours or send me email. I will usually need to see some kind of SPSS output, code or error statements in order to answer your questions - please bring them along or include them in your email.



Quizzes: There will be two in-class quizzes of about 45 minutes. The quizzes will include multiple choice, short answer and computational exercises. You may use your text books, notes and other written references on the quizzes.



Exams: The midterm and final exams will be take home tests. In addition to the references permitted on the quizzes, you may also come to me with your questions. However, unlike the homework, you are not permitted to share answers with other students on the take-home exams.



Other stuff: I recommend that you have a calculator that can (at least) add, subtract, multiply, divide, calculate squares and square roots, and save the sum of a sequence of numbers.



If you plan to use statistics in your own research and you have a computer at home, you should at least consider purchasing the graduate student edition of SPSS for Windows. It's a great buy at around $165 - approximately 85% less than the regular price! If you need an additional standard reference on SPSS, one that many people like is:

Norusis, Marija J. SPSS Guide to Data Analysis. Englewood Cliffs, NJ: Prentice Hall.