Sociology 536 – Quantitative Methods in Sociology II

Spring 2003

INSTRUCTOR:

Paul D. Allison, 276 McNeil, 215-898-6717, www.ssc.upenn.edu/~allison, allison@ssc.upenn.edu

Ofc. Hours: Tues. & Thurs. 1:30-2:30

TEACHING ASSISTANT:

Adair Crosley, 254A McNeil, (215) 573-8061, acrosley@ssc.upenn.edu

LECTURES: Tues. & Thurs., 12:00 - 1:30

CONTENT: This is the second semester of a two-semester sequence. It is a course on applied linear models, with an emphasis on multiple regression and its extensions, including path analysis and simultaneous equations. Students will use the SAS statistical package.

TEXTS:     1. McKee J. McClendon, Multiple Regression and Causal Analysis

        2. Paul D. Allison, Multiple Regression: A Primer

        3. Lecture notes (available at Campus Copy, 3907 Walnut).
 

EXAMS: A midterm and a final exam. Dates will be announced at least two weeks in advance

RECITATIONS:  Thurs. 2-3, 4-5.  Sessions will typically include review and explanation of lecture material, explanation of difficult homework problems, and responses to students' questions on any of the course material. New material will be presented on the use of the computer.

WRITTEN ASSIGNMENTS: There will be several written assignments at irregular intervals throughout the semester. Most of these will require you to analyze a set of data using SAS.

GRADING: Final grades will be primarily determined by exam performance, with the midterm accounting for 40 percent and the final accounting for 60 percent. Poor performance on problem sets can detract from your grade, however.

PREREQUISITE: Sociology 535 or equivalent course in introductory statistics.

TENTATIVE SEQUENCE OF TOPICS:

1. Bivariate regression algebra.

2. Bivariate regression theory.

3. Trivariate and multiple regression.

4. Statistical inference in multiple regression.

5. Interpretation of regression coefficients.

6. Regression and correlation; standardized coefficients.

7. Multicollinearity.

8. Nonlinearity.

9. Dummy variables, analysis of variance and analysis of covariance.

10. Interaction.

11. Model building strategies.

12. Missing data.

13. Violation of assumptions.

14. Path analysis.

15. Generalized least squares.

16. Fixed-effects and random-effects.

17. Simultaneous equations.