Multiple hypothesis testing in experimental economics

By incorporating information about dependence ignored in classical multiple testing procedures, such as the Bonferroni and Holm corrections, the bootstrap-based procedure has much greater ability to detect truly false null hypotheses.