1GVPT 722: Quantitative Methods for Political Science IISpring 20089:30-12:15 Monday (Patterson 1229)Professor: Geoff Layman TA: Ozan KalkanOffice: Tydings 2126C Office: Tydings 2126Office Hours: 1:30-3:30 Mon/Wed, and Office Hours: 3:00-5:00 Tues., 2:00-4:00 by appointment Thurs., and by appointmentOffice Phone: 5-9709 Office Phone: 5-6350E-Mail: [email protected] E-Mail: [email protected] Webpage: http://www.bsos.umd.edu/gvpt/glayman/html/gvpt_722.htmlCourse DescriptionThis course is the second part of the two-semester sequence in quantitative methods. In GVPT622, we discussed the fundamentals of Ordinary Least Squares (OLS) regression, and learned thatOLS is the best method of estimating the statistical relationship between two or more variableswhen certain assumptions hold. However, when those assumptions do not hold, OLS loses manyof its desirable statistical properties, and other methods of estimation become more appropriate. The focus of this course will be on the nature of the OLS assumptions, the circumstances underwhich they do and do not hold, and extending the general linear model to estimate relationshipsamong variables when the assumptions of OLS do not hold.While we will devote a good bit of attention to the mathematical and theoretical foundations ofvarious statistical methods, the primary goal of the class is to enable you to apply these methodsto substantive research questions. To that end, we will pay particular attention to estimating andinterpreting models with political science data, and you will produce several small papers in whichyou will do just that.Statistical Software and Computer LabsWe will use the statistical software program Stata for homework assignments and for empiricalexamples in class. Stata is available for your use in both the government and politics computer labin Tydings 1115 and in the OACS labs in Lefrak Hall. You also may purchase a copy of Stata forhome use at http://www.stata.com/order/new/edu/gradplans/gp-direct.html. Either the $95version (Stata/IC 10, one-year license) or $155 version (Stata/IC 10, perpetual license) will befine for the purposes of this class. If you have not had experience in using Stata, you should go to the course webpage anddownload the introduction to Stata handout and work your way through the first few chapters ofthe Pollock book. Chapter 2 of the Long and Freese book is also very useful for this. You alsomight want to set up an appointment with Ozan so that he can provide a brief tutorial.2Texts All of these books should be available for purchase at the University Book Center and MarylandBook Exchange. The following book is the main text for the course. I highly recommendpurchasing your own copy. • Damodar N. Gujarati. 2003. Basic Econometrics, 4th ed. New York: McGraw Hill.We will read about half of the following book, and Professor Hanmer covers nearly the entirebook in his course on maximum likelihood estimation. I have placed an earlier edition of the bookon reserve in McKeldin Library if you do not wish to purchase it.• J. Scott Long and Jeremy Freese. 2006. Regression Models for Categorical DependentVariables Using Stata, 2nd ed. College Station, TX: Stata Press.The following books are ones that we either will read a good bit of or are part of therecommended reading. I will place copies of all of them in 3140 Tydings. You may copy themthere if you would prefer not to purchase them. Please do not take them out of the main office. These books also can be ordered directly from the Sage publications website (http://www.sagepub.com/series.aspx?sid=486&sc=2).• Melissa A. Hardy. 1993. Regression with Dummy Variables. Newbury Park, CA: Sage.• Herbert B. Asher. 1983. Causal Modeling, 2nd ed. Newbury Park, CA: Sage.• Christopher H. Achen. 1982. Interpreting and Using Regression. Newbury Park, CA:Sage.• James Jaccard, Robert Turrisi, and Choi K. Wan. 2003. Interaction Effects in MultipleRegression, 2nd ed. Newbury Park, CA: Sage.We read most of the following book in 622. If you did not take 622 from me, you may considerpurchasing it for an introduction to Stata and access to data sets in Stata format.• Phillip H. Pollock III. 2006. A Stata Companion to Political Analysis. Washington, D.C.:CQ Press.Course Requirements and Grade Distribution1. Homework and Quizzes (20%): Nearly every week, you will be asked to do a short homeworkassignment to turn in at the beginning of the next class. Most of the homework assignments willask you to calculate statistics by hand and interpret their meaning, but some may ask you toconduct short data analyses with Stata and interpret the results, and others may be moreconceptual. We also may have a few short quizzes (both announced and unannounced) over thecourse of the semester to assess your understanding of the material. Some of the questions forquizzes may be computational, so you should always bring a calculator to class.32. Data Analysis Papers (12% each, 48% total): You will be asked to write four short papers applying the methods learned in the class to substantive research questions using political sciencedata. I will provide you with the data and with instructions as to the type of analyses to conductfor each paper. You will conduct the analyses and write a short paper in which you explain theanalysis, present its results, and interpret those results. The topics of the papers and their duedates are not yet set in stone, but here are the likely topics and the approximate due dates. It ispossible that we will fall behind and not be able to cover enough material for you to do the lastpaper in a timely fashion. If that circumstance arises, you will only do three papers and each ofthem will be worth 16% of the course grade.Topic Approximate Due DateMultiple Regression and the Assumptions of OLS March 3Regression with Interaction Terms March 31Regression with a Dichotomous Dependent Variable April 21Regression with Panel Data May 123. Final Exam (32%): There will be a cumulative final exam given in class at our normal classtime (9:30 a.m.) on May 19. The exam will include conceptual questions about the nature andapplicability of particular statistical methods, questions that ask you to compute statisticalestimates, and questions that ask you to interpret statistical estimates.Course Policies1. Homework assignments and papers should be turned in at the beginning of class on their duedates. Late homeworks and papers will be