Goals of the Course:

First, to provide an introduction to the methods of event history analysis so that students with a reasonable background in quantitative methods can use the methods on their own empirical problem. Second, to introduce more quantitatively sophisticated students to the literature, so that they can continue reading on their own. Finally, to provide a forum in which people with event history data can explore different ideas and approaches to analyzing that data. There will be one exam on the material at the end of the technical part of the course, and a required paper. Grading will be based on the test (40%) and the paper (40%), and class participation (20%).
 

Textbooks:

Allison, Paul D. (a) Survival Analysis Using the SAS System: A Practical Guide. SAS Institute 1995.

Allison, Paul D. (b) Event History Analysis: Regression for Longitudinal Event Data. Sage 1982.

Yamaguchi, Kazuo Event History Analysis Sage 1991.

Everyone will also read parts of several other books and journal articles, some of which are below.
 

Additional Required Papers:

Hannan, Michael T. "Macrosociological Applications of Event History Analysis: State Transitions." Quality and Quantity 1989. pp.351-383.

Hannan, Michael T. and Tuma, Nancy Brandon "Methods for Temporal Analysis." Annual Review of Sociology; 1979, 5, 303-328.

Kiefer, N.M. 1988. "Economic Duration Data and Hazard Functions." Journal of Economic Literature. XXVI(June 1988). Pp.646-679.

Olzak, Susan "Analysis of Events in the Study of Collective Action." Annual Review of Sociology 1989, 15, 119-141.
 

Books suggested, but not required:

Blossfeld, Hamerle and Mayer Event History Analysis Lawrence Erlbaum 1989.

Tuma and Hannan Social Dynamics: Models and Methods Academic Press 1984.

Very useful resources:

Cox, D. R. And D. Oakes. 1984. Analysis of Survival Data. New York: Chapman and Hall.

Hosmer and Lemeshow. 1992. Logistic Regression.

Kalbfleisch, J.D. and R. L. Prentice. 1980. The Statistical Analysis of Failure Time Data. New York: Wiley.
 

Outline of Topics:

I. Introduction (Yamaguchi ch. 1, Allison (a) 1-14, Allison (b) 7-14, Blossfeld ch. 1,2, Tuma and Hannan ch. 1,2)

A. What is event history analysis?

1. The analysis of categorical variables in continuous time.

2. The analysis of transitions

3. The regression analysis of durations

B. Why Event History Analysis?

1. The creation and destruction of social entities

2. The equilibrium assumption

3. The availability of data

C. What do we need to know first?

 
1. Linear models: specification and estimation

2. Math: algebra, very simple calculus, exponents

3. Curves: exponential, logarithmic

4. How to run SAS programs on the mainframe: (LIFETEST, LIFEREG, PHREG, LOGISTIC)

D. What kinds of dynamic models are there?

1. Review of linear models

a. Specification error
b. Measurement error
c. Autocorrelation
d. Causal lag

2. Discrete time models

3. Continuous time models

E. What is distinctive about event history data?

1. Censoring

2. Repeatable and non-repeatable events

3. Competing risks

II. Basic concepts in EHA (Hannan 1989, Tuma and Hannan ch. 3, Allison (a) 15-28, Allison (b) 22-32, Hannan 1989)

A. Data structures in EHA: time, state spaces, censoring

1. Cross section

2. Panel

3. Event sequence

4. Event count

5. Backward recurrence times

6. Event history

B. The survivor function

C. The hazard function

D. The density function

E. Time dependence in parametric models: exponential, Weibull, Gompertz.

III. Estimating event history models:

A. Describing Event History Data:

1. The Kaplan-Meier Estimator (Allison (a) 29-41, Tuma and Hannan 43-57)

2. The life-table method (Allison (a) 41-51)

B The maximum likelihood principle.( B ch 4,Tuma and Hannan 116-128)

C. Parametric Analysis. (Allison (a) Ch. 4, Allison (b) 22-29, Kiefer, Hannan 1989).

D. Partial Likelihood. (Allison (a) Ch. 5, Yamaguchi chapters 5&6, Blossfeld ch 5, Allison (b) 33-85)

E. Competing risks. (Allison (a) ch 7).

F. The discrete-time approximation

1 The logistic regression model (Yamaguchi chapter 1, Allison (a) ch 7, Allison(b) 1-22, Hosmer and Lemeshow ch 1-3)

2. Discrete approximation in event history analysis (Yamaguchi chapter 2)

G. Other issues. (Allison (a) ch 8)

IV. Event history data management.

A. Constructing Event History Data

1. Archival Sources

2. The Life History Calendar

B. Managing Event History Data: SAS as a database manager.
 

V. Student projects.

A. Topic presentation and discussion.

B. Data management issues.

C. Analysis strategies and results.

D. Data presentation and writing strategies.
 
 

After the introduction to event history methods is completed, we will begin reading event history applications. An abstracted list of references to choose from is available, and will be circulated. Each student will choose a reference from the list (or elsewhere, with approval), and present a review of the material in class.