On the Treatment of Grouped Observations in Life Studies.
Abstract
Cox (Regression models and life tables, J. Roy. Statist. Soc. 34B, 1972) presents a systematic study of the use of covariates in the analysis of life time. Cox's basic model is that of proportional failure rates. A particularly thorny point in theory and application is the treatment of tied observations. It appears that grouping rather than discrete time is the proper way to handle ties. This paper studies what to do with grouped observations. Maximum likelihood leads to ignoring intervals in which no failures occur. A logistic model, which makes explicit use of Cox's earlier binary data methods, is introduced and illustrated with a numerical example. When the grouping becomes fine this logistic model leads back to the proportional failure rate model suggested by Cox. (Author)
Document Details
- Document Type
- Technical Report
- Publication Date
- Apr 15, 1976
- Accession Number
- ADA024436
Entities
People
- W. A. Thompson Jr.
Organizations
- University of Missouri