a bootstrapping procedure in which samples are randomly drawn (with replacement) from an observed data set, values for different population characteristics are estimated, and the estimates averaged across the set of samples. Then, the samples created during this initial procedure are themselves sampled and their parameters estimated and averaged. The resampling and recalculating process continues until a stable parameter estimate is obtained. Iterated bootstrapping usually is less prone to error than is the use of a single set of bootstrapping samples.