What Hypothesis Tests are Not: a Response to Colegrave and Ruxton

It is always flattering to see one’s work cited by others. Not only does it boost the ego, but it provides a satisfying feeling that one’s efforts are both appreciated and contributing to the advance of science. So I was pleased when a colleague pointed out that Colegrave and Ruxton (2003) had cited a recent paper of mine, ‘‘The insignificance of statistical significance testing’’ ( Johnson, 1999). In that article I argued, as did Colegrave and Ruxton, that confidence intervals often are much more informative than are p values associated with hypothesis tests