Biometrika · 1973 · 528 citations · 3 references
Public HealthBiostatisticsEpidemiologic MethodsStatistical ModelingSurvival AnalysisHazard FunctionsLife Table MethodsParametric Survival ModelsSurvival FunctionsSurvival ModelingTime-to-event AnalysisCox RegressionFailure-time ModelsLikelihood In Survival AnalysisMarginal Survival ModelsMonte Carlo EstimationProbabilistic Survival ModelsSurvival Risk EstimationSurvival Risk Modeling
Marginal likelihoods are obtained for the regression parameters in the model presented by Cox (1972). If no ties occur in the recording of failure time data the results of Cox are given a straightforward justification. If ties occur in the data, results different from those suggested by Cox are obtained. Some Monte-Carlo comparisons of these competing results are made. A discrete model is developed for grouped data from Cox's model and estimates of the survivor function are given for both continuous and grouped data.
3