Applied Regression The University of Texas at Dallas EPPS 6316, Spring 2013 Tuesday, 7pm 9:45pm Room: FO 2.410 Professor: J.C. Barnes, Ph.D. Email: jcbarnes@utdallas.edu Phone: (972) 883-2046 Office: GR 2.206 Office hours: Tuesday 2 4pm and by appointment Course Description and Learning Outcomes This course provides a survey of bivariate and multivariate regression techniques. In that sense, this is an intermediate-level course, where the student is expected to have a firm grasp on algebra and basic probability theory/application. The focus of this course is primarily on the Ordinary Least Squares (OLS) regression model. This application focused course presents examples drawn from economics, political science, public policy, sociology, and criminology. Basic concepts germane to econometric/statistical analysis will be introduced first, followed by a review of bivariate analytic techniques such as correlation. Finally, the OLS model will be introduced with attention being given to the estimation of parameter values, interpreting parameter estimates, and diagnosing problems with the OLS model. An introduction to the statistical package, Stata, will also be provided as the course unfolds. All analytic procedures introduced in this course will be reinforced with a demonstration of the method in Stata. The goal of this course is to provide the student with a solid foundation and understanding of basic regression estimation techniques en route to a conceptual understanding of advanced statistical techniques. Students are expected to have mastered the following by the end of this course: 1) appropriate application of the OLS model (knowing when and where regression can be applied); 2) understand the assumptions of the OLS model (what conditions must be met in order for the OLS model to be applied and the results be meaningful); 3) be able to diagnose problems with the OLS model (recognize when an OLS assumption is violated and how to diagnose it); 4) be able to identify corrective measures (know how to modify the analysis when assumptions are violated); and 5) interpretation of findings (be able to interpret the substantive meaning of OLS regression results). Required Reading 1) Allison, P. (1999). Multiple Regression: A Primer. Pine Forge Press. ISBN: 0761985336 2) Gordon, R. A. (2010). Regression Analysis for the Social Sciences. Routledge. ISBN: 9780415991544
3) Articles posted to elearning (see Course Schedule below) Course Requirements & Grading Your grade will be determined based on your performance on four (4) assignments, your performance on two (2) examinations, and your performance on one (1) empirical research paper. The first assignment will be worth 5 points to your final grade. The remaining three assignments will be worth 10 points to your final grade. You will have one (1) week to complete each assignment. More details about the assignments will be provided in class as their due dates approach. The midterm will count for 20 percent of your final grade and the final exam will count for 20 percent of your final grade. The exams will consist of a mixture of multiple choice and short answer questions. The remaining 25 percent of your final grade will be gleaned from your performance on an empirical research paper. More details about the research paper will be provided in class as the due date approaches. Your final grade will be determined using the following rubric and grading scale: Percent of Total Item 5 Assignment 1 10 Assignment 2 10 Assignment 3 10 Assignment 4 20 Midterm Exam 20 Final Exam 25 Empirical Research Paper Grading Scale A = 93-100 A- = 90-92 B+ = 87-89 B = 83-86 B- = 80-82 C+ = 77-79 C = 70-76 F = 0-69 Course Policies Attendance is required if you wish to receive a passing grade. If you miss a class, I strongly encourage you to borrow one of your classmate s notes and then talk with me if you need additional clarification. I do not provide students with my notes nor will I post notes/slides to the Internet.
If you miss an exam or an assignment you must: (1) notify me within 24 hours AND (2) provide me with an acceptable excuse. Note that I may request written documentation. If you do not follow this procedure you will receive a zero on the examination/assignment. Technical Support If you experience any problems with your UTD account you may send an email to: assist@utdallas.edu or call the UTD Computer Helpdesk at 972-883-2911. University Policies http://go.utdallas.edu/syllabus-policies
Course Schedule and Assigned Readings (A=Allison text; G=Gordon text) Lecture Topic Stata Topic Due Reading 1/15 Course introduction -- -- -- 1/22 Identifying datasets, univariate stats basic command logic, generating vars. -- -- & introduction to Stata reading in data, codebook 1/29 Intro to data analysis recode, generate, summary, tabulate Assign. 1 G: 1, 2, 3, 4 (levels of measurement, remedial bivariate stats) 2/5 Intro to data analysis t-test, ANOVA, χ 2 -- A: 5; G: 5 (remedial bivariate stats) 2/12 Bivariate regression I correlate, bivariate regression estimation Assign. 2 A: 5; G: 5 2/19 Significance testing & probability theory -- -- -- 2/26 Multivariate regression I multivariate regression estimation Assign. 3 A: 1; G: 6 (Concepts, algebra, basic estimation, residuals) 3/5 Midterm Exam -- -- -- 3/12 Spring Break -- -- -- 3/19 Multivariate regression I, continued multivariate regression estimation -- A: 2; G: 6 (Concepts, algebra, basic estimation, residuals) 3/26 Multivariate regression II multivariate regression estimation -- A: 4, 7; (dummy variables and multicollinearity) G: 7, 11.3
4/2 Multivariate regression III polynomials and interactions Assign. 4 A: 8; G: 8, 9 (non-linear effects) (non-additivity; Clogg test) 4/9 Multivariate regression IV heteroscedasticity, autocorrelation, -- A: 3, 6; G: 11 (OLS assumptions and diagnostics) non-normality, outliers, leverage 4/16 Multivariate regression V Sobel mediation, detecting spuriousnes -- G: 10 (direct, indirect, and spurious effects) 4/23 Multivariate regression VI logistic, poisson, nbreg, tobit -- A: 9 (generalized linear models) 4/30 Final Exam -- -- -- 5/10 No in-class meeting -- Empirical Research Paper -- *Posted to elearning Note: This schedule is not a binding contract. I reserve the right to make changes at any time and it is your job to stay abreast of these changes.