# Statistical Advisor, Analyzing Failure Times

Use SURVIVAL ANALYSIS. This chapter discusses various analytic routines commonly used with censored or uncensored survival/failure time data.

A data set is said to be 'censored' if some observations are incomplete, but not missing. For example, in industrial reliability testing, a part or product may not fail (break) within the time span covered by a study. However, we do not know how long it will function properly thereafter, and thus, that observation is 'censored.'

SURVIVAL ANALYSIS allows you to describe the distribution of failure times via life tables analysis or the Kaplan-Meier product limit method. Various theoretical distributions can be fit to the data (Weibull, Gompertz, exponential, linear hazard). That module also contains several tests for comparing failure times in two or more groups or samples. It also allows you to evaluate the relationship of one or more continuous predictor variables to the observed failure times. Various linear and nonlinear regression models are supported (normal and lognormal regression, exponential regression, proportional hazard regression).

A key aspect of product quality is product __reliability__. A number of specialized techniques are available to quantify reliability and to estimate the "life expectancy" of a product. The methods are discussed in *Weibull analysis & reliabilty/failure time analysis*.