Analysis of doubly censored Burr type II distribution: a Bayesian look

Abstract

Burr model is especially suitable for the life-testing of the products that age with time. Trimmed samples are widely utilized in several areas of statistical practice, especially when some sample values at either or both extremes might have been adulterated. In this article, the problem of estimating the parameter of Burr distribution type II based on trimmed samples under informative and uninformative has been addressed. The problem discussed using Bayesian approach to estimate the shape parameter of Burr type II distribution. Elicitation of hyperparameter through prior predictive approach has also been discussed. Posterior predictive distributions along with posterior predictive intervals and credible intervals have also been derived under different priors. A comparison has been made using the Monte Carlo simulation. A real life data example has also been discussed.