Author/Authors :
başar, emel gazi üniversitesi - fen fakültesi - istatistik bölümü, Ankara, Turkey
Title Of Article :
Gamma Frailty Models in Survival Analysis
Abstract :
Most of the models used in the analysis of survival data have beenassumed that all individuals had the same risk of failure. This situation is notalways realistic because some of individuals have different failure structure thanthe others. This heterogeneity in the population could cause serious problem andcould lead to misleading conclusions. To describe the unobserved heterogeneity insurvival data frailty models are used. The frailty factor is usually assumed to actmultiplicatively on a baseline hazard function. Frailty models are conditionalmodels and the frailty factor is random. Therefore a frailty distribution needs to bespecified in the frailty model. In this study gamma frailty model has beenconsidered under the proportional hazards assumption and the models have beenillustrated with real data set.
NaturalLanguageKeyword :
Survival analysis , Proportional hazards model , Gamma frailty model , Parametric models , Child delinquency
JournalTitle :
Journal Of Natural and Applied Sciences
