We found a match
Your institution may have access to this item. Find your institution then sign in to continue.
- Title
Duration analysis in longitudinal studies with intermittent observation times and losses to followup.
- Authors
Hajducek, D. M.; Lawless, J. F.
- Abstract
We consider the analysis of spell durations observed in event history studies where members of the study panel are seen intermittently. Challenges for analysis arise because losses to followup are frequently related to previous event history, and spells typically overlap more than one observation period. We provide methods of estimation based on inverse probability of censoring weighting for parametric and semiparametric Cox regression models. Selection of panel members through a complex survey design is also addressed, and the methods are illustrated in an analysis of jobless spell durations based on data from the Statistics Canada Survey of Labour and Income Dynamics. The Canadian Journal of Statistics 40: 1-21; 2012 © 2012 Statistical Society of Canada
- Subjects
CANADA; REGRESSION analysis; STATISTICS; INCOME; PROBABILITY theory
- Publication
Canadian Journal of Statistics, 2012, Vol 40, Issue 1, p1
- ISSN
0319-5724
- Publication type
Article
- DOI
10.1002/cjs.10139