Portal de Periódicos da CAPES

Interval Censoring

2005; Linguagem: Inglês

10.1002/0470011815.b2a11039

Jianguo Sun,

Statistical Methods in Clinical Trials

Abstract In interval censoring, a random variable is known only to lie in an interval. In survival or failure‐time data, this may arise, for instance, when a critical change has occurred sometime between a patient's successive visits to a clinic. The survival distribution may be estimated by nonparametric maximum likelihood. Attention is given to the estimation of parameters in regression models and to the comparison of survival functions.