Advances in computer technology are making it possible for educational researchers to use simpler statistical methods to address a wide range of questions with smaller data sets and fewer, and less restrictive, assumptions. This digest introduces computationally intensive statistics, collectively called resampling techniques. Resampling is a process for estimating probabilities by conducting vast numbers of numerical experiments with the aid of high speed computers. Probability theory has freed researchers from the drudgery of repeated experiments, and resampling has now made it possible to analyze virtually any statistic in an easy to understand manner, without making assumptions about the distribution of the data. The following four techniques are important in resampling: (1) the bootstrap; (2) the jackknife; (3) cross-validation; and (4) balanced repeated replication. An example compares the grades of voucher and non-voucher students in an elementary school using the bootstrap. It is argued that resampling frees researchers from the assumption that the data conform to a bell-shaped curve and the need to focus on statistical measures with theoretical properties that can be analyzed mathematically. Critics question the resampling method itself and question the accuracy of the estimates that resampling yields. In some situations, resampling may be less accurate than conventional parametric methods. (SLD)