The bootstrap is a computer intensive method in Statistics, which was introduced by Efron in 1979, and can be used to estimate different aspects of accuracy of a sampling distribution like bias, standard deviation, construct confidence intervals and more. The bootstrap is known to work in many examples, however, there are examples in which it fails. In such cases, modifying the bootstrap sample size can rectify the problem. The question raised is which sample size to use.

In the first part of the talk I will introduce the bootstrap method and in the second part present an adaptive rule to choose the bootstrap sample size.