Bootstrap sample size crucial for accurate extreme value predictions.
The study found that when using bootstrapping to estimate extreme values or intermediate quantiles from a sample, the size of the bootstrap sample needs to be smaller than the original sample size. This means that in order to accurately predict extreme values or specific quantiles, a smaller subset of the data is needed for the bootstrapping process. This finding aligns with previous research and highlights the importance of considering sample size when using bootstrapping in extreme value theory.