Sociology 229A:  Event History Analysis

 

Spring 2008, Class #69745

 

Time/Place:

Tuesday 2:00-4:50pm, SSPB 4206

Class Web Page:

http://www.socsci.uci.edu/~schofer/2008Soc229EHA/home229EHA.htm

 

 

Instructor:

Evan Schofer

Office:

SSPB 4271

Office Hours

Tuesday/Thursday 12:45-1:45 and by appointment

Office Phone:

(949) 824-1397

Email:

 

 

Introduction

 

Event history analysis (EHA), also called survival analysis or failure analysis, refers to a set of statistical tools for describing and modeling processes that unfold over time.  EHA is useful for studying the occurrence of time-specific events, such as the birth of a child, the termination of a war, the incidence of protests, the creation of strategic alliances among firms, and so on.  We will examine the most common types of event history techniques and models, including non-parametric tools (e.g., graphs, tests of inequality of survivor functions), semi-parametric models (e.g., Cox Regression), and fully parametric event history models (e.g., exponential or Weibull models), as well as diagnostics.  The course will also cover basic issues, such as the nature and organization of event history datasets and the interpretation of EHA results.  We will be using the statistical software package Stata. 

 

 

Readings

 

Readings can be found in the main texts or online via the UCI e-reserve website.  Additionally, you may receive handouts and other small reading assignments on occasion.  Complete reading assignments prior to the class in which material will be covered.  You will get much more out of class if you have already finished the readings. 

 

Main Texts:

 

Box-Steffensmeier, Janet M. and Bradford Jones.  2004.  Event History Modeling:  A Guide for Social Scientists.  Cambridge, UK:  Cambridge University Press.

 

Cleves, Mario, William W. Gould, and Roberto Gutierrez.  2004.  An Introduction to Survival Analysis Using Stata, Revised Edition.  Stata Press.

 

Required Online Readings:

 

You can also the UCI E-Reserves system via the following URL, and then searching by course name or instructor:

http://antpac.lib.uci.edu/search/r

Or, use the following direct link:

http://antpac.lib.uci.edu/search?/pSchofer/pschofer/-3%2C-1%2C0%2CB/frameset&FF=pschofer&2%2C%2C2

 

IMPORTANT INFORMATION ABOUT E-RESERVE ACCESS:

 

Details about how to access E-Reserves can be found here:

http://www.lib.uci.edu/services/reserves/res_eres.html

 

Important excerpts from that web page:

What is needed to access E-Reserves?

You must have a valid UCI Library card. You will also need a Library PIN (Personal Identification Number) to use E-Reserves. If you do not yet have a valid UCI Libraries card, you will need to activate your card in person at the Loan Desk at Langson, Science or Grunigen Medical Library. Valid identification and proof of current UCI enrollment or employment is required to activate your library card.

If you are using E-Reserves from off-campus, you will need to be authenticated as a current UCI student, faculty or staff member in order to access licensed electronic materials available through E-Reserves. Please refer to Connecting from Off-Campus.

Note the latter:  Authentication is required to access readings from off-campus locations.  Depending on your computer setup, this might involve entering your id and password in a web browser or installing and using VPN software.  If you can’t do this, you must print or download the readings from a computer while you are on-campus.

 

 

Stata Software

 

We will be using the statistical software package Stata.  It is available on machines in the computer labs in the Social Science Tower (SST 170 and SST 604).  Both labs are open during the day.  Nighttime access to the grad student lab (SST 604) requires a key, which can be obtained from Mario Nunez.  You may also purchase Stata for your personal computer via NACS at the link below.  Costs range from $50 for a limited version (sufficient for this class, but not for projects with large datasets) to $150 for a permanent license to “intercooled Stata” (which can handle large datasets).  I recommend the latter if you can afford it.

http://licenses.nacs.uci.edu/detail2.cfm?ProductID=57

 

 

Assignments and Evaluation

 

Mini-Assignments.  There will be several short assignments, collectively worth 75% of your final grade.

 

Short paper.  There will be a short paper worth 25% of your final grade.

 

Assignments received late will be marked down one partial grade (i.e., and A becomes an A-, C+ becomes a C) per day past the due date.  Extensions will be granted for legitimate reasons if requested in advance – before the due date. 

 

Your final grade will be computed based on the percentage weightings indicated.  In the event of a borderline grade, I may use my discretion in adjusting grades based on course participation, improvement, and effort (or lack thereof).  Incompletes will not be given, except in unusual circumstances.

 

 

General Information

 

Check the course web site periodically.  Urgent notices may be posted on the web site (e.g., if an assignment due date were to be extended).  Also, the course web page will contain important information:  copies of course handouts and assignments, data relevant to course assignments, etc. 

 

 

Schedule & Reading Assignments

 

Week 1:  Introduction and Review  (April 1)

 

Start on readings for next week!

 

 

Week 2:  Basic Concepts and Terminology  (April 8)

 

Box-Steffensmeier and Jones, Chapter 1 “Event History and Social Science.” 

Box-Steffensmeier and Jones, Chapter 2 “The Logic of Event History Analysis.”

Cleves, Gould, and Gutierrez, Chapter 1 “The Problem of Survival Analysis.”

 

Long, J. Scott and Jeremy Freese.  2006.  “Introduction to Stata.”  Chapter 2 in Regression Models for Categorical Dependent Variables Using Stata (Second Edition).  College Station, TX:  Stata Press.  Reading available on the UCI E-Reserve system.  (Optional reading for those needing a Stata refresher.) 

 

Additional helpful Stata information can be found here:  http://www.cpc.unc.edu/services/computer/presentations/statatutorial

 

 

Week 3:  Non-Parametric Approaches:  Graphs & Tests  (April 15)

 

Cleves, Gould, and Gutierrez, Chapter 2 (focus on section 2.3), “Describing the Distribution of Failure Times.” 

Cleves, Gould, and Gutierrez, Chapter 4 “Censoring and Truncation.” 

Cleves, Gould, and Gutierrez, Chapter 5 “Recording Survival Data.” 

Cleves, Gould, and Gutierrez, Chapter 8 “Nonparametric Analysis.” 

 

Blossfeld, Hans-Peter, Karin Golsch, and Gotz Rohwer (eds.).  2007.  “Event History Data Structures.”  Chapter 2 in Event History Analysis with Stata.  New York:  Lawrence Erlbaum Associates, Inc.  (Optional reading)  UCI E-Reserve.

 

 

Week 4:  Event History Models  (April 22)

 

Cleves, Gould, and Gutierrez, Chapters 3.  “Hazard Models.”

Box-Steffensmeier and Jones, Chapter 4 “The Cox Proportional Hazards Model.”

Box-Steffensmeier and Jones, Chapter 5 “Models for Discrete Data.”

 

Empirical Example:  Long, J. Scott, Paul D. Allison, and Robert McGinnis.  1993.  “Rank Advancement in Academic Careers:  Sex Differences and the Effects of Productivity.”  American Sociological Review, 58, 5:703-722.  UCI E-Reserve.

Also at: http://links.jstor.org/sici?sici=0003-1224%28199310%2958%3A5%3C703%3ARAIACS%3E2.0.CO%3B2-F

 

 

Week 5:  Event History Model (cont’d)  (April 29)

 

Cleves, Gould, and Gutierrez, Chapter 9 (section 9.1 only), “The Cox Proportional Hazards Model.”

Cleves, Gould, and Gutierrez, Chapter 10, “Model Building Using stcox.” 

Cleves, Gould, and Gutierrez, Chapter 11 (section 11.1 only).  “The Cox Model:  Diagnostics.”

 

Empirical Example:  Hironaka, Ann M.  2005.  “World Patterns in Civil War Duration.”  Chapter 2 in Neverending Wars.  Cambridge, MA:  Harvard University Press.  UCI E-Reserve.

 

 

Week 6:  Event History Models and Diagnostics  (May 6)

 

Cleves, Gould, and Gutierrez, Chapter 12 (focus on section 12.1) “Parametric Models.”

Cleves, Gould, and Gutierrez, Chapter 13 (focus on 13.0, 13.1.1, 13.2.1) “A Survey of Parametric Regression Models in Stata.” 

Box-Steffensmeier and Jones, Chapter 3 “Parametric Models for Single-Spell Duration Data” and

 

Empirical Example:  Schofer, Evan.  2003.  “The Global Institutionalization of Geological Science, 1800-1990.”  American Sociological Review, 68 (Dec): 730-759.  UCI E-Reserve.

Also at: http://links.jstor.org/sici?sici=0003-1224%28200310%2968%3A5%3C730%3ATGIOGS%3E2.0.CO%3B2-C

 

 

Week 7:  Event History Models and Diagnostics  (May 13)

 

Cleves, Gould, and Gutierrez, Chapter 11 (section 11.2 only) “The Cox Model:  Diagnostics.”

Cleves, Gould, and Gutierrez, Chapter 14 (skim) “Post-estimation Commands for Parametric Models.”

Box-Steffensmeier and Jones, Chapter 8 “Diagnostic Methods for Event History Analysis.” 

 

Empirical Example:  Soule, Sarah A and Susan Olzak.  2004.  “When Do Movements Matter? The Politics of Contingency and the Equal Rights Amendment.”  American Sociological Review, Vol. 69, No. 4. (Aug., 2004), pp. 473-497.  UCI E-Reserve.

Also at:  http://links.jstor.org/sici?sici=0003-1224%28200408%2969%3A4%3C473%3AWDMMTP%3E2.0.CO%3B2-F

 

 

Week 8:  Problems and Issues in Event History Analysis (May 20)

 

Box-Steffensmeier and Jones, Chapter 6 “Issues in Model Selection.”

Box-Steffensmeier and Jones, Chapter 11 “The Social Sciences and Event History.”

 

Blossfeld, Hans-Peter, Karin Golsch, and Gotz Rohwer (eds.).  2007.  “Introduction.”  Chapter 1 in Event History Analysis with Stata.  New York:  Lawrence Erlbaum Associates, Inc.   UCI E-Reserve.

 

Blossfeld, Hans-Peter, Karin Golsch, and Gotz Rohwer (eds.).  2007.  “Problems of Model Specification.”  Chapter 10 in Event History Analysis with Stata.  New York:  Lawrence Erlbaum Associates, Inc.  UCI E-Reserve.

 

 

Week 9:  Advanced Topics  (May 27)

 

Cleves, Gould, and Gutierrez, Chapter 15 “Generalizing the Parametric Regression Model.” 

Cleves, Gould, and Gutierrez, Chapter 9 (section 9.4 only) “The Cox Proportional Hazards Model.” 

Box-Steffensmeier and Jones, Chapter 9 “Some Modeling Strategies for Unobserved Heterogeneity.”

 

Additional Readings:  To Be Announced, if time allows. Otherwise, wrap up older topics.

 

 

Week 10:  Advanced Topics  (June 3)

 

Additional Readings:  To Be Announced, if time allows. Otherwise, wrap up older topics.

 

June 3:  Short Paper Due