ABSTRACT
Statisticians often complain that researchers will come along with a set of data and ask for advice as to how to analyze it. In his Presidential Address to the First Indian Statistical Congress in 1938, Sir Ronald Fisher commented: “To consult the statistician after an experiment is finished is often merely to ask him to conduct a post mortem examination. He can perhaps say what the experiment died of.” P-values are very often misunderstood, and there are plenty of instances of wrong definitions being given even in textbooks. There are many criticisms of the use of p-values in science, and a good case can be made for using alternative approaches, notably methods based on Bayes theorem. The author's focus here is on Null Hypothesis Significance Testing in part because such a high proportion of studies in the literature use this approach, and it is important to understand p-values in order to evaluate the literature.
