%0 Journal Article
%T Comparison of Bayesian and Classical Analysis of Weibull Regression Model: A Simulation Study
%A £¿mran KURT £¿M¨¹RL¨¹
%A Mevl¨¹t T¨¹RE
%A Kaz£¿m £¿ZDAMAR
%J Turkiye Klinikleri Journal of Biostatistics
%D 2011
%I Turkiye Klinikleri
%X Objective: The purpose of this study was to compare performances of classical Weibull Regression Model (WRM) and Bayesian-WRM under varying conditions using Monte Carlo simulations. Material and Methods: It was simulated the generated data by running for each of classical WRM and Bayesian-WRM under varying informative priors and sample sizes using our simulation algorithm. In simulation studies, n=50, 100 and 250 were for sample sizes, and informative prior values using a normal prior distribution with was selected for b1. For each situation, 1000 simulations were performed. Results: Bayesian-WRM with proper informative prior showed a good performance with too little bias. It was found out that bias of Bayesian-WRM increased while priors were becoming distant from reliability in all sample sizes. Furthermore, Bayesian-WRM obtained predictions with more little standard error than the classical WRM in both of small and big samples in the light of proper priors. Conclusion: In this simulation study, Bayesian-WRM showed better performance than classical method, when subjective data analysis performed by considering of expert opinions and historical knowledge about parameters. Consequently, Bayesian-WRM should be preferred in existence of reliable informative priors, in the contrast cases, classical WRM should be preferred.
%K Bayes theorem
%K survival analysis
%K markov chains
%K computer simulation
%U http://www.turkiyeklinikleri.com/pdf/?pdf=4cfb52a46bbfa4b5f77f5ec24aa8de03