‘Statistical computing’ explores the role of information technology in statistical analysis. As computer power has increased, models have become more and more advanced. Somewhat counterintuitively, this can cause them to be a worse fit with the wider population. Cross validation splits the sample in half, and uses one of these halves to evaluate the model. Bootstrap sampling is another approach which estimates the uncertainty in complex models. Computing power has revolutionized both Bayesian statistics and statistical graphics. The development of dedicated statistical packages has given non-experts access to advanced tools. However, statistical knowledge is still required to apply these techniques correctly.