ENBIS-13 in Ankara

15 – 19 September 2013 Abstract submission: 5 February – 5 June 2013

Weibull Analysis with Heavily Censored Data

18 September 2013, 10:15 – 10:35

Abstract

Submitted by
Nikolaus Haselgruber
Authors
Nikolaus Haselgruber (CIS Consulting in Industrial Statistics)
Abstract
The Weibull distribution is a frequently preferred model in the analysis of lifetime data of technical products. In this context, the fitting of the distribution to data by parameter estimation is usually referred to as Weibull Analysis. If this is applied to warranty data, the problem of heavy censoring is faced as such data contain information excusively about the warranty period which covers in general only the initial phase of the useful life period. In addition, the proportion of failing products is small compared to the population.
This presentation discusses the effects of heavily censored data on the parameter estimation of the Weibull distribution. It should provide a guidline about what aspects to consider when the question about lifetime estimation based on warranty data is raised.
View paper

Return to programme