Parameter Estimation for Weibull Burr Type X Model with Right Censored Data

Abstract

Studies have considered generalizing statistical distributions in the past. These were aimed at making such distributions more flexible and suitable for describing real-world phenomena. In this study, we considered exploring the Weibull Burr Type X distribution, which extends the Burr Type X distribution using the Weibull generator. Particularly, the performance of the maximum likelihood estimators for its parameters encompassing the right censored dataset was explored and compared. On the performance of its estimators with respect to bias and root mean square error, we considered the Monte Carlo simulation study to make a comparison using varying sample sizes and censored percentages. We illustrated the usefulness and potentials of the Weibull Burr Type X distribution using a right censored dataset. We considered comparing the fitness of this model to its sub-models using real world dataset. The result showed that the Weibull Burr Type X distribution provides a better fit than other competing models. This indicates that the distribution is flexible and competitive. The Weibull Burr Type X distribution exhibits unimodal and decreasing shapes. The extra parameter in the distribution varies the model's tail weight and introduces skewness into the model. We introduced this model as an alternative to other existing models for modelling right censored data in various research fields and areas of study.