An Application of a Zero‐inflated Lifetime Distribution with Multiple and Incomplete Data Sources

We analyze data sampled from a population of parts in which an associated anomaly can occur at assembly or after assembly. Using a zero-inflated lifetime distribution to fit left-censored and right-censored data as well data from a supplementary sample, we make predictions about the proportion of the population with anomalies today and in the future. Goodness-of-fit is also addressed. Copyright (C) 2016 John Wiley & Sons, Ltd.