The Jackknife and Bootstrap in Biomedical Research—Common Principles and Possible Pitfalls

Abstract

This article briefly reviews the basic principles of two resampling techniques, namely the jackknife and the bootstrap. Small examples are given to address possible failures of these methods. Their application is finally illustrated by two substantial problems, that is, the estimation of a common odds ratio in stratified contingency table analysis and the statistical evaluation of individual bioequivalence. The former is used to demonstrate the benefit of the jackknife in terms of bias reduction and the latter to exhibit the sensitivity of the bootstrap to minor changes in the bootstrap algorithm.