Topics in Methodology

• Sign up for the class mailing list (it's working now).
• I'm trying to reserve time for the poster session. Either Friday, April 30 (10-12) or Monday, May 3 (11-2).
• The monte carlo projects are due to me by the end of exam week (May 14).
• The poster session is set for April 30th from 10:00 to 12:00 in The Commmons. See the information on how to do a good poster on the Department's Resources for Graduate Students web page and make sure to review the requirements in the class syllabus.
• Remember to drop off your Monte Carlo projects during exam week and to set up an appointment to talk about it with me.

Due February 19, 2004: Recreate Figure 1 in Gelpi and Griesdorf, but based on a multinomial logit rather than an ordered probit model.

Due February 23, 2004: Prepare a proposal for your research project for this class. Treat this is as if you were submitting a proposal to a conference. Your proposals must be circulated to the class email list by 5pm so that we can read them before class the next day.

Due March 03, 2004: Write a monte carlo program that generates y as a function of two correlated independent variables, then (incorrectly) estimate a heteroskedastic probit with one of the variables in the variance term but not in the index. Vary the correlation from -0.9 to 0.9 and graph the average value of the two coefficients. Take a look at the week 5 computer files below for a head start.

1. Discrete Choice Analysis.
2. Survival Analysis.
3. Event Count Analysis.
4. Monte Carlo.

• UCLA Stata tutorial.
• Political Methodology Working Paper Archive
• Political Methodology Video Series
• Stata tutorial from UCLA
• An archive of Gauss resources
• Gary King's software page
• ICPSR data archive
• Richard Tucker's Homepage
• A horribly brief list of Stata commands, written for the intro class.

• Course syllabus.
• Monte Carlo examples from Week 1.
• Discrete choice Monte Carlo from Week 2.
• Heteroskedastic Probit Monte Carlo from week 5.
• Illustration of Clarify and outreg from week 6.
• Poisson data. for week 7.
• Negative binomial program and data. for week 8.
• Truncated Poisson and NB models. for week 10.
• Discrete EHA Data. for week 13.
• EHA data with time-varying and time-invariant covariates for week 14. Here is the Stata batch file to analyze the former data set.
• Monte Carlo program for ignoring duration dependence for Week 15.
• Data for running duration models with selection for Week 16.

Week 1: Course Introduction, Monte Carlo Analysis.
Reading: King 1990; Beck 1999; and Mooney & Krause 1997.
Week 2: Binary Choice Models.
Reading: Long, Ch. 3-4.
Week 3: Ordered Logit and Probit.
Reading: Long, Ch. 5; Gelpi & Griesdorf 2001; Tolbert, McNeal and Smith 2003.
Week 4: Multinomial Logit and Probit.
Reading: Long, Ch. 6; Alvarez and Nagler 1998.
Week 5: Heteroskedastic Logit and Probit.
Reading: Alvarez and Brehm 1995.
Week 6: Research Proposal Presentations.
Reading: Nagler 1995 (JSTOR download); King, Tomz and Wittenberg 2000 (JSTOR download).
Week 7: Poisson regression model.
Reading: Long, Ch. 8.1-8.2; King 1988.
Week 8: Negative binomial and GEC models.
Reading: Long, Ch. 8.3; King 1989a.
Week 9: Spring Break.
Week 10: Hurdle, Zero-Inflated and Truncated Count Models
Reading: Long, Ch. 8.4-8.7; King 1989c.
Week 11: Applications of Count Models.
• Reading for Tuesday: Wahlbeck, Spriggs and Maltzman 1998; Shields and Huang 1995.
• Reading for Thursday: Krain 1997; Bercovitch and Schneider 2000.
Week 12: Introduction to Event History.
• Reading: Box-Steffensmeier and Jones 1997; Beck 1997; Box-Steffensmeier and Jones 2004, Ch 1,5.
Week 13: Example: State Policy Adoption.
• Reading: Box-Steffensmeier and Jones 2004: Ch 3.
Week 14: The Cox Model, Time-Varying Covariates.
• Reading: Box-Steffensmeier and Jones 2004: Ch 4,6,7.
Week 15: Repeated Failures, Competing Risks.
• Reading: Box-Steffensmeier and Jones 2004: Ch 10,11; Box-Steffensmeier and Zorn 2002.
Week 16: Miscellaneous Topics.
• Reading: Boehmke, Morey and Shannon 2004 (Download); Alt, King and Signorino 2001.