Statistical Hypothesis Testing in Biology: A Contradiction in Terms

Abstract

We respond to the preceding article by Perry (J. Econ. Entomol. 79: 1149–1155) concerning the subject of hypothesis testing and meaningful presentation of statistical analyses. We agree with Perry that indiscriminate use of statistical hypothesis testing should be discouraged. We also address and refute reasons often given to support continued emphasis on hypothesis testing. We conclude that biologists should take the responsibility for evaluating whether or not the magnitude of the difference between population parameters or treatment effects is biologically important, instead of merely testing for whether such a difference exists. We recommend that authors display the estimate of the difference and the confidence limit for this difference. These statistics, along with the standard error, contain the most information: their presentation provides the clearest and most meaningful format of the statistical analyses.