ABSTRACT
Analysis of product lifetime data typically requires concepts and tools different from those taught in most introductory courses on statistics. For example, lifetime data often involve censored observations on units whose exact failure times are unknown. In response, appropriate models and methods for proper analysis of lifetime data have been developed by statisticians. In this chapter we review and illustrate some important concepts and statistical distributions that are commonly used in lifetime data analysis of products and components that are non-repairable and whose life, therefore, ends at the time of failure. These include:
Reliability function.
Hazard function.
Weibull, exponential and lognormal distributions.
Censoring of lifetime data.
Probability plotting.
Key assumptions of lifetime data analysis.
Dangers of extrapolation.
We also offer a review of software for product lifetime data analysis.
