Ziliak and McClosky's criticisms of significance tests: A damage assessment

摘要: D. N. McCloskey and Stephen Ziliak have criticized economists others for confounding statistical substantive significance, committing the logical error of transposed conditional. In doing so they sometimes misinterpret function significance tests. Nonetheless, make both these errors ? but not nearly as often claim. They also argue ?incorrectly that existence an effect, which is what tests are about, a scientific question. Their complaint in testing do take loss into account unfounded. But right arguing confidence intervals should be presented more frequently.