| Time: | 12:30 pm on Wednesday, October 28, 2009 |
| Place: | Savery 409 |
The challenges of conducting cohort analysis are well known. Whether any observed time-related differences from data are related to aging or cohort variation or historical time is a question usually deemed conceptually important but empirically intractable. What should one do to model these effects to understand the social and biological mechanisms generating the data? I aim to provide some useful guidelines on how to conduct cohort analysis. I first summarize the state of knowledge on cohort analysis in sociology, demography, and epidemiology. I then provide a synthesis of new developments, including models, methods, and substantive applications, for three commonly used research designs: 1) age-by-period tables of rates; 2) repeated cross-sectional surveys; and 3) accelerated longitudinal panels. I finally discuss avenues for future research.
To request disability accommodations, contact the Office of the ADA Coordinator ten days in advance of the event. 543-6450 (voice) 685-3885 (FAX) 543-6452 (TDD) access@u.washington.edu