Robust estimation based on one-shot device test data under log-normal lifetimes

TL;DR

This paper develops robust estimators for one-shot device test data assuming log-normal lifetimes, along with confidence intervals and Wald-type tests, validated through simulations and numerical examples.

Contribution

It introduces robust estimation methods specifically designed for one-shot device data with log-normal lifetime assumptions, including inference tools.

Findings

Robust estimators outperform traditional methods in contaminated data scenarios.

Confidence intervals and tests maintain accuracy under model deviations.

Simulation studies demonstrate improved robustness and reliability.

Abstract

In this paper we present robust estimators for one-shot device test data under lognormal lifetimes. Based on these estimators, confidence intervals and Wald-type tests are also developed. Their robustness feature is illustrated through a simulation study and two numerical examples.