Abstract:
The problem of number of parametric bootstrap replications that are necessary to realizace at bias estimation is solved in this paper. It shows results of simulations demonstrating extent of error that we can cause at different number of bootstrap replications.