Research Methods II (SYA 6306)

Spring 2008

Professor: Chris Girard

W 6:25 pm – 9:05 pm

http://www.fiu.edu/~girardc

e-mail: girardc@fiu.edu

cell: 305-244-4668

Office hours: MWF 10-11 A.M.

Web resources for learning Stata-9:

http://www.ats.ucla/edu/stat/seminars

http://www.ats.ucla/edu/stat/

Course content:

This is the second course in a two-semester sequence of courses on social research methods. The first-semester course focused on descriptive and inferential statistics leading up to linear regression. This second course will focus on multiple regression, a technique frequently used by sociologists for examining the effect of several explanatory (independent) variables on a dependent variable. This technique allows the analyst to assess the effect of each variable, controlling for the effects of the others, as well as examine joint contributions. One of the major objectives of the course is to be able to produce a paper suitable for publication in a social science journal.

Prerequisite: SYA 6305 (“Research Methods I”) or an equivalent course.

Required books:

Mendenhall & Sincich, A Second Course in Statistics: Regression Analysis, 6th ed. (Prentice Hall, 2003);

Allison, Multiple Regression: A Primer (Pine Forge, 1999);

Pyrczak, Fred. Evaluating Research in Acamdemic Journals. Fourth edition. (Pyrczak Publishing, 2008).

There will also be a handout from the following book: Hamilton, Lawrence C., Regression with Graphics Pacific Groves, CA: Brooks/Cole, 1992. (chapter 7 Logit Regression).

Highly recommended is Dummeldinger, Student’s Solution Manual (which supplements Mendenhall & Sincich); and Hamilton, Statistics with Stata – Updates for Version 9.

Required software

The previous semester’s materials presented an introduction to Stata, the statistical software program that will be integral to this course. Students will be expected to have routine access to a professional version of Stata-9.

We will also learn how to use the program via the free, downloadable resources for learning Stata (as well as SPSS, SAS, and other statistical programs) that are available at the following web sites of UCLA-Academic Technology Services:

http://www.ats.ucla.edu/stat/

http://www.ats.ucla.edu/stat/seminars/

http://www.ats.ucla.edu/stat/stata/

http://www.ats.ucla.edu/stat/stata/webbooks/reg/chapter1/statareg1.htm

See the modules Learning Stata I & II and Training Videos for Stata. Specifically important for this course is the four-chapter “web book” (with practice data) on using Stata-8 for regression analysis:

http://www.ats.ucla.edu/stat/stata/webbooks/reg/default.htm

The course’s objective, however, is learning regression analysis and solving practical problems that arise in the course of data analysis. Learning statistical software is subordinate to this pursuit.

Homework, Projects, Exams & Grades

Final grades will be based on the following:

Weekly homework problems. These will be graded pass/fail and

collectively worth 25% of the final grade. Passing all such assignments earns an “A,” all but one of them a “B,” and so on. Homework problems assigned at a given class session are due at the start of the next class session. Germane Stata output must be edited and pasted into a word document with your own interpretation inserted in appropriate places.

One-project. It will be worth 25% of the final grade. The project will entail writing a methods and results section of a paper that presumably could be submitted for publication in a journal. The paper will contain an appendix (not normally included in a publishable manuscript) containing all diagnostic tests conducted, an interpretation of these test results, and any raw output from Stata used to generate tables in the results section. Every student will select in consultation with the instructor a data set to analyze for the projects

Two exams. Each of these will be worth 25% of the final grade.

Schedule:

Jan. 9: What is regression analysis?

Mendenhall & Sincich, chapter 2, “Introduction to Regression Analysis”

Mendenhall & Sincich, chapter 3, “Simple Linear Regression”

Recommended: Mendenhall & Sincich, chapter 1, “A Review of Basic

Concepts”; Allison, chapter 5, “Bivariate Regression”;

Assignment: (1) Go to the Stata website (by clicking the link on the Stata program) and subscribe to the Stata listserv digest. (2) Select an article using OLS regression from either American Sociological Review, American Journal of Sociology, or Social Forces since the year 2000. Briefly summarize how the research problem is conceptualized and how the data are described and analyzed. Then copy (by pen or pencil) the format of the main tables that describe the data and report the regression results. (3) Use StatTransfer to transfer the data from one of the Mendenhall & Sincich problems into Stata. (4) Mendenhall & Sincich problems 3.11, 3.25, 3.32, 3.45, 3.49, 3.63, 3.75.

Jan. 16, 23: Multiple regression

Allison, chapter 1, “What Is Multiple Regression?”

Allison, chapter 2, “How Do I Interpret Multiple Regression Results?” Mendenhall & Sincich, chapter 4, “Multiple Regression”

Recommended: Chen et al., Regression with Stata, chapter 1, “Simple and

Multiple Regression” http://www.ats.ucla.edu/stat/stata/webbooks/reg/default.htm

Assignment: (1) Mendenall & Sincich problems 4.1, 4.11, 4.13, 4.19, 4.27, 4.43, 4.45 4.65, 4.79, 4.81

Jan. 30 & Feb. 6: Model building

Allison, chapter 3, “What Can Go Wrong with Multiple Regression?”

Allison, chapter 8, “How Can Multiple Regression Handle Nonlinear Relationships?”

Mendenhall & Sincich, chapter 5, “Model Building”

Chen et al., Regression with Stata, chapter 3, “Regression with Categorical

Predictors” http://www.ats.ucla.edu/stat/stata/webbooks/reg/default.htm

Pyrczak, Chapters 1-2

Assignment: (1) Mendenhall & Sincich problems 5.3, 5.15, 5.21, 5.25, 5.29, 5.35. 5.43; Pyrczak, exercises for Chapters 1-2

Feb. 13: Selecting variables

Mendenhall & Sincich, chapter 6, “Variable Screening Methods”

Pyrczak, Chapters 3-4

Assignment: (1) Mendenhall & Sincich problems 6.1, 6.3, 6.5;

Pyrczak, exercises in Chapters 3-4

Feb. 20: Regression diagnostics, part 1

Allison, chapter 6, “What Are the Assumptions of Multiple Regression?”; and chapter 7, “What Can Be Done about Multicollinearity?”

Mendenhall & Sincich, chapter 7, “Some Regression Pitfalls”

Chen et al., Regression with Stata, chapter 2, “Regression Diagnostics”

http://www.ats.ucla.edu/stat/stata/webbooks/reg/default.htm

Pyrczak, Chapters 5-6

Assignment: (1) Mendenhall & Sincich problems 7.5, 7.7, 7.11, 7.17, 721;

Pyrczak, exercises in Chapters 5-6

Feb. 27: Regression diagnostics, part 2

Mendenhall & Sincich, chapter 8, “Residual Analysis”

Pyrczak, Chapters 7-8

Assignment: (1) Mendenhall & Sincich problems 8.1, 8.5, 8.17, 8.25, 8.31, 8.37, 8.39, 8.41;

exercises in Pyrczak, Chapters 7-8

Mar. 5: MID-TERM EXAM (call 305 244 4668 before exam to schedule

make-up; otherwise you get an automatic 0)

Mar. 12 & 26: Regression with categorical dependent variables—binary dependent variables

Hamilton, Regression with Graphics (chapter 7 Logit Regression).

Chen et al., Regression with Stata, chapter 4, “Beyond OLS”

http://www.ats.ucla.edu/stat/stata/webbooks/reg/default.htm

Pyrczak, Chapters 9-10

Assignment: (1) Select an article using logistic, ordinal, or multinomial regression from either American Sociological Review, American Journal of Sociology, or Social Forces since the year 2000. Briefly summarize how the research problem is conceptualized and how the data are described and analyzed. Then copy (by pen or pencil) the format of the main tables that describe the data and report the regression results. (2) Mendenhall & Sincich – chapter 9.6;

exercises in Pyrczak, Chapters 9-10.

Apr. 2, 9, 16: Regression with categorical dependent variables—

ordinal & nominal dependent variables

Chen et al., Regression with Stata, chapter 4, “Beyond OLS”

http://www.ats.ucla.edu/stat/stata/webbooks/reg/default.htm

Pyrczak, Chapters 11-13

Assignment: exercises in Pyrczak, Chapters 11-13

Project rough draft Due April 9

project DUE APRIL 16

FINAL EXAM: To be Announced