PARAMETRIC DURATION MODELS FOR CENSORED AND UNCENSORED DATA
No Thumbnail Available
Date
2017
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
Ahmadu Bello University Zaria, Nigeria
Abstract
Censoring is said to be present in data when information on time to outcome event is not available for all the study participants. The purpose of this study was to compare some parametric survival models when data are censored and when uncensored. The study used duration of labour data obtained from Record and files Department, Federal Medical Centre, Lokoja. Parametric proportional hazards model, namely, Exponential, Weibull and Gompertz models were fitted when the data were censored and when uncensored. Only variable full housewife was found to be significant in predicting duration of labour. Also, it was found that the three models had better fits when the data were censored than when they were uncensored. The study concludes that it is better to include censored observation when analysing duration data despite the incomplete information they provide.
Description
Keywords
Censoring, proportional hazard, duration of labour, AIC, uncensored.