This work presents a bayesian inference on discrete Weibull distribution with censored data. A Full Bayesian Significance Test (FBST) was proposed to test the shape parameter of model. Samples from the posterior distributions of parameters were numerically obtained by Markov Chain Monte Carlo (MCMC) simulations. This methodology was illustrated using simulated data and by application on a real database of survival times of men diagnosed with AIDS. All simulations and estimates were performed in R language.
survival analysis; bayesian inference; hypothesis tests; MCMC