DESCRIPTION

This course is designed to help you develop a solid working knowledge of both event history analysis and panel data models. It would be ideal for students considering a dissertation using such techniques, but also appropriate for those who just want more hands-on experience with methodologies that have become increasingly popular in both scientific and policy work. Event methods are terrific when researchers want to predict whether and when something happens -- for example, wars, births, deaths, strikes, crimes, or job promotions. Using examples that take nations, states, and individuals as the units of analysis, we'll cover topics such as demographic life tables, survival and hazard analysis, competing risks, proportional hazards, and time-varying covariates. In the second half of the course, we'll bridge from the concept of time-varying predictors to panel models in which both the independent and dependent variables are changing over time. Here we’ll cover lagged dependent variables, first differences, fixed and random effects, clustering, and other topics. The course will be pitched at a level that will make it accessible to anyone who has taken the sociology graduate statistics course (8811) or equivalent. We’ll focus on the basics and learning to apply these basics well.

OBJECTIVES

1. The course will help you develop a more nuanced understanding of two important classes of quantitative research techniques in the social sciences. We’ll emphasize useful and powerful general tools, such as proportional hazards models for the analysis of events and fixed effects models for the analysis of panel data.

2. The seminar will provide support, encouragement, and an opportunity for you to make significant progress on one or more of your own research projects.

3. In using and explaining these models, you will sharpen your written and oral research presentation skills.

4. We will work through empirical pieces by some top sociological researchers. As you develop your own research, it is useful to see how others have translated propositions into testable hypotheses, devised appropriate methodologies to test them, and presented the results to diverse audiences.

5. The course will stimulate your thinking about research design and analysis more generally. This includes how we produce our knowledge, its meaning and relevance, and the utility of various tools. Such big-picture considerations may help you to make creative research choices that are personally meaningful and professionally rewarding. I’ll encourage you to articulate and develop your own research values, principles, and orientation to the field.

6. Finally, a graduate seminar should encourage your professional development as you make the transition from student to independent social scientist. I will share "backstage" information from the articles we read, including files, data sets, and other materials that may show you different facets of the research and publication process.

TEXTS

Allison, Paul D. Fixed Effects Regression Models. 2009. Thousand Oaks, CA: Sage. [$17.55]

Box-Steffensmeier, Janet M. and Bradford S. Jones. 2004. Event History Modeling: A Guide for Social Scientists. Cambridge: Cambridge University Press. [$30.15]

You will also read some challenging research articles throughout the semester, but I’ve limited the number of required readings to just a few per week – emphasizing solid and accessible applications of the techniques we discuss, rather than the brilliant statistical innovations that brought us these techniques. I’ve put some of my own work on this syllabus –not because it is exemplary but to help us talk through concrete design considerations and research choices.  

RECOMMENDED REFERENCES

We will dip into the following books this term, so you will get a chance to preview many different approaches to the subject. Those interested in developing practical expertise in the area should purchase and reference them (if not read them, stem to stern).

• Allison, Paul D. 1984. Event History Analysis: Regression for Longitudinal Event Data. Thousand Oaks: Sage. [I’ve found all of Allison’s books useful].
• Blossfeld, Hans-Peter, Katrin Golsch, and Gotz Rohwer. 2009 [2007]. Event History Analysis with Stata. New York: Taylor and Francis (Psychology Press).
• Cleves, Mario, Roberto Gutierrez, William Gould, and Yulia Marchenko. 2010. An Introduction to Survival Analysis Using Stata, 3rd Ed. College Station, TX: Stata Press.
• Cox. D. R., and D. Oakes. 1984. Analysis of Survival Data. London: Chapman and Hall.
• Finkel, Steven E. 1995. Causal Analysis with Panel Data. Thousand Oaks, CA: Sage.
• Hamilton, Lawrence C. 2013. Statistics with Stata, Version 12. [for those new to Stata]
• Sophia Rabe-Hesketh and Anders Skrondal. 2012. Multilevel and Longitudinal Modeling Using Stata (3d). College Station, TX: Stata Press.
• Sayrs, Lois W. 1989. Pooled Time Series Analysis. 1989. Thousand Oaks, CA: Sage.
• Singer, Judith D. and John B. Willett. 2003. Applied Longitudinal Data Analysis: Modeling Change and Event Occurrence. New York: Oxford University Press.
• Yamaguchi, Kazuo. 1991. Event History Analysis. Newbury Park, CA: Sage.

TENTATIVE OUTLINE AND REQUIRED READINGS

1. 1/23 WELCOME and THE BIG PICTURE

“Every designers’ dirty little secret is that they copy other designers’ work. They see work they like, and they imitate it. Rather cheekily, they call this inspiration.” — Aaron Russell

Introductions; Course logic; Requirements; Software

Keywords
Change; Events; Causality; Design; Becoming a responsible “user"

2. 1/30 EVENT HISTORY BASICS & NON-PARAMETRIC APPROACHES [628 OPEN]

“A designer can mull over complicated designs for months. Then suddenly the simple, elegant, beautiful solution occurs … it feels as if God is talking! And maybe He is.” – Leo Frankowski

Keywords
Censoring and Truncation; Survival; Hazard distributions; Life tables; Kaplan-Meier

How To
Box-Steffensmeier and Jones. Chapters 1 and 2.  Pp. 1-20.

Blossfeld, Hans-Peter, Katrin Golsch, and Gotz Rohwer. 2009 [2007]. Event History Analysis with Stata. New York: Taylor and Francis (Psychology Press). Pp. 58-86. Chapter 3. Nonparametric Descriptive Methods.

Hamilton, Lawrence C. 2013. Statistics with Stata, Version 12. Chapter 10, Survival and Event Count Models, Pp. 283-293. [especially useful for those unfamiliar with Stata]

A Clear Example
Elizabeth Arias. 2010. United States Life Tables by Hispanic Origin. National Center for Health Statistics. Vital and Health Statistics, Series 2(#152).

My Reference Point [not required]
Christopher Uggen, Jeff Manza, and Melissa Thompson, 2006. “Citizenship, Democracy, and the Civic Reintegration of Criminal Offenders.” The Annals of the American Academy of Political and Social Science 605:281-310. [see methods appendix]

3. 2/6 PARAMETRIC AND SEMI-PARAMETRIC MODELS

“Design is a plan for arranging elements in such a way as best to accomplish a particular purpose.” — Charles Eames

Keywords
Specifying baseline hazard (exponential, Weibull, log logistic, log normal, Gompertz);
Assessing fit; Proportional hazards

How To
Box-Steffensmeier and Jones. Chapter 3.  Pp. 31-46. Parametric Models for Single-Spell Durations.
Box-Steffensmeier and Jones. Chapter 4.  Pp. 47-68. The Cox Proportional Hazards Model.

A Clear Example (or Two)
Peter S. Bearman and Hannah Brückner. 2001.”Promising the Future: Virginity Pledges and First Intercourse.” American Journal of Sociology 106:859-912.

Daniel J. Myers. 1997. “Racial Rioting in the 1960s: An Event History Analysis of Local Conditions.” American Sociological Review 62:94-112.

My Reference Point [not required]
Christopher Uggen. 2000. "Work as a Turning Point in the Life Course of Criminals: A Duration Model of Age, Employment, and Recidivism." American Sociological Review 65:529-46.

4.  2/13 A WHOLE FREAKING DAY ON DATA AND DESIGN [628 OPEN]

“Data is [are] sexy.” — Hans Rosling

        Due: two-page brainstorming proposal, specifying the following:

• Research questions
  • Offer both an event history question and a panel data question
  • e.g., Does inter-ethnic economic competition predict genocide? [an event]
  • e.g., Does inter-ethnic economic competition predict inter-ethnic homicide? [changes in a level or rate]
• Your data possibilities and needs
  • (1) Currently do-able; (2) do-able with effort; or (3) maybe after tenure?
  • e.g., I’ve got a list of genocides with dates and nations that I could connect with a publicly available 100-nation data set. [I’d call this a “2”]
• Analytic approach
  • How will your design answer the question?
  • e.g., I should be able to do some basic non-parametric analysis and test bivariate relationships – I’ve got to think harder about ways to isolate the effects of economic competition from competing explanations. [Fair enough]
• Likely Barriers
  • How far do you think you can get this semester?
  • e.g., I’d like to create basic survival and hazard plots and run some basic multivariate models, but I may not be able to chase down all the controls I’ll need in 10 weeks.
  • What’s your back-up plan for completing presentations and the seminar project?

Discussion: Setting up and analyzing person-period data

Keywords
Person-period data

5. 2/20 MULTIPLE EVENTS, COMPETING RISKS and MODEL SELECTION

“Truly elegant design incorporates top-notch functionality into a simple, uncluttered form.” — David Lewis

Keywords
Repeated events; Competing risks; Royston-Parmar

How To
Box-Steffensmeier and Jones, Chapter 6. Issues in Model Selection Pp. 85-94.
Box-Steffensmeier and Jones, Chapter 10. Models for Multiple Events Pp. 155-82.

A Clear Example (or Two)
Palloni, Alberto, Douglas S. Massey, Miguel Ceballos, Kristin Espinosa, and Michael Spittel. 2001. “Social Capital and International Migration: A test Using Information on Family Networks.” American Journal of Sociology 106: 1262–98. [multistate hazard example]

Michelle Budig. 2006. “Intersections on the Road to Self-Employment: Gender, Family, and Occupational Class.” Social Forces 84:2223-39. [competing risk example]

My Reference Point [not required]
Candace Kruttschnitt, Christopher Uggen, and Kelly Shelton. 2000. "Predictors of Desistance among Sex Offenders: The Interaction of Formal and Informal Social Controls." Justice Quarterly 17:61-87. [competing risk example]

6. 2/27 TIME-VARYING COVARIATES / DISCRETE-TIME LOGITS [628 OPEN]

“Design is easy. All you do is stare at the screen until drops of blood form on your forehead.” — Marty Neumeier

Keywords
discrete-time; TVC (15);

How To
Box-Steffensmeier and Jones, Chapter 5. Models for Discrete Data Pp. 69-83.
Box-Steffensmeier and Jones, Chapter 7.  Inclusion of Time-Varying Covariates Pp. 95-117.

A Clear Example
Daniel Schneider. 2011. “Wealth and the Marital Divide.” American Journal of Sociology 117:627-67. [discrete time]

My Reference Point [not required]
Angela Behrens, Christopher Uggen, and Jeff Manza. 2003. “Ballot Manipulation and the ‘Menace of Negro Domination’: Racial Threat and Felon Disenfranchisement in the United States, 1850-2002.” American Journal of Sociology 109:559-605.

7. 3/6  FROM TIME-VARYING COVARIATES TO PANEL DATA

“Our opportunity, as designers, is to learn how to handle the complexity, rather than shy away from it, and to realize that the big art of design is to make complicated things simple.” — Tim Parsey

Keywords
Regression review; Static-score; lags; differencing; GMM

How To
Steven Finkel, 1995. Causal Analysis with Panel Data. Chapter 2, Modeling Change with Panel Data. Pp. 2-21.
 
A Clear Example (or two)
Guang Guo and Leah Vanwey.  “Sibship Size and Intellectual Development:  Is the Relationship  Causal?” 1999.  American Sociological Review 64:169-87. [see also comments and replies 188-206]

Wesley Longhofer and Evan Schofer. 2011. “National and Global Origins of Environmental Association.” American Sociological Review 75:505-33. [event and dynamic panel]

My Reference Points [not required]
Heather McLaughlin, Christopher Uggen, and Amy Blackstone. 2012. “Sexual Harassment, Workplace Authority, and the Paradox of Power." American Sociological Review 77:625-47.

Michael Massoglia and Christopher Uggen. 2010. “Settling Down and Aging Out: Toward an Interactionist Theory of Desistance and the Transition to Adulthood.” American Journal of Sociology116:543-82. [Simple lags]

8. 3/13 CATCH-UP AND PRESENTATIONS [628 OPEN]

“Every designers’ dirty little secret is that they copy other designers’ work. They see work they like, and they imitate it. Rather cheekily, they call this inspiration.” — Aaron Russell

Due: slides for 5-10-minute talk and write-up (no more than 5 pages):

• EHA research question(s)
• Data
• Results: Non-parametric, “bivariate,” and multivariate
• Caveats and next steps to improve the analysis

3/20 – NO CLASS – SPRING BREAK

9. 3/27  LINEAR FIXED EFFECTS  MODELS [628 OPEN]

“The most innovative designers consciously reject the standard option box and cultivate an appetite for thinking wrong.” — Marty Neumeier

Keywords
LSDV; xtreg

How To
Allison, Chapters 1-2. Linear Fixed Effects Models. Pp. 1-26.

A Clear Example
Waldfogel, Jane. 1997. “The Effect of Children on Women’s Wages.” American Sociological Review 62:209–17.

My Reference Point [not required]
Christopher Uggen and Melissa Thompson. 2003. "The Socioeconomic Determinants of Ill-Gotten Gains: Within-Person Changes in Drug Use and Illegal Earnings." American Journal of Sociology 109:146-85. [fixed effects and differencing]

10.  4/3 FIXED EFFECTS MODELS II – TIME SERIES APPROACHES AND EVENTS

Keywords
Stationarity, balance

How To
Nathaniel Beck and Jonathan N. Katz. 2011. “Modeling Dynamics in Time-Series Cross-Section Political Economy Data.” Annual Review of Political Science 14:331-52.

A Clear Example (or two)
David Jacobs and Jason T. Carmichael. 2001. “The Politics of Punishment across Time and Space: A Pooled Time-Series Analysis of Imprisonment Rates.” Social Forces 80:61-89.

[RECOMMENDED ONLY] Mauro F. Guillen and Sandra L. Suarez. 2005. “Explaining the Global Digital Divide: Economic, Political and Sociological Drivers of Cross-National Internet Use.” Social Forces 84:681-708.

My Reference Point [not required]
The Contingent Effect of Incarceration on State Health Outcomes.

11. 4/10  FIXED EFFECTS LOGIT, COUNT, AND EVENT MODELS [628 OPEN]

“Design is where science and art break even.” — Robin Mathew

Keywords
GEE, Poisson, Negative Binomial

How To
Allison, Chapter 3. Fixed Effects Logistic Models, Pp. 28-48.
Allison, Chapter 4. Fixed Effects for Count Data, Pp. 49-69.
Allison, Chapter 5. Fixed Effects for (Repeatable) Events, Pp. 70-86.

A Clear Example
Bushway, Shawn, Robert Brame, and Raymond Paternoster. 1999. “Assessing Stability and Change in Criminal Offending: A Comparison of Random Effects, Semiparametric, and Fixed Effects Modeling Strategies.” Journal of Quantitative Criminology 15:23–61.

12. 4/17  BRIDGING FROM FIXED EFFECTS TO STRUCTURAL EQUATIONS

“Design should never say, “Look at me.” It should always say, ‘Look at this.’” — David Craib

Keywords
Instrumental variables; MPlus; Latent variables

How To
Allison, Chapter 6. Structural Equation Models with Fixed Effects, Pp. 87-98.

A Clear Example
Kenneth A. Bollen and Jennie E. Brand. 2010. “A General Panel Model with Random and Fixed Effects: A Structural Equation Approach.” Social Forces: 89:1-34.

13.  4/24  BRIDGING TO MIXED AND MULTI-LEVEL APPROACHES [628 OPEN]

“Design is the application of intent - the opposite of happenstance, and an antidote to accident.” — Robert L. Peters

Kewords
Growth curve; Latent trajectory; Level-1 and Level-2

How To
Singer, Judith D. and John B. Willett. 2003. Applied Longitudinal Data Analysis: Modeling Change and Event Occurrence. New York: Oxford University Press. Chapter 3. Introducing the Multilevel Model for Change. Pp. 45-74.

Sophia Rabe-Hesketh and Anders Skrondal. 2012. Multilevel and Longitudinal Modeling Using Stata (3d). College Station, TX: Stata Press.  Chapter 7. Growth-Curve Models. Pp. 343-82.

Hamilton, Lawrence C. 2013. Statistics with Stata, Version 12. Chapter 13, Multilevel and Mixed-Effects Modeling, Pp. 387-421.

14. 5/1  CATCH-UP AND PRESENTATIONS

“Good design is a renaissance attitude that combines technology, cognitive science, human need, and beauty to produce something that the world didn’t know it was missing.” — Paola Antonelli

        Due: slides for 5-10-minute talk and write-up (no more than 5 pages):

• Longitudinal research question(s)
• Data
• Results: present a basic cross-sectional model (e.g., OLS regression), a lagged dependent variable or difference model, and a fixed or random effect model
• Caveats and next steps to improve the analysis

15. 5/8 LAST DAY - LET'S DO LUNCH!

“Design is not the narrow application of formal skills, it is a way of thinking.” — Chris Pullman

Due: projects

 

REFERENCE LINKS

Want More? It was difficult to keep the reading list manageable, but I’ve got lots more recommendations for classic treatments and contemporary applications of these techniques.

HOME
CV 1-PG
Google Page

ADVISEES

RESEARCH
Writing Grants
Felon Voting

TEACHING
4111 4141
8001 8111 8890

OUTREACH
Blog Media
Links Quotes
TheSocietyPages