Hypothesis testing and p-values

In 1925 Ronald Fisher's book Statistical Methods for Research Workers suggested that a result with a probability below 1 in 20 under the null hypothesis was worth taking seriously. He meant it as a convenient guide. Within decades, p < 0.05 had become a gate for publication in many fields.

In 2016 the American Statistical Association published a formal statement on p-values, the first time it had taken an official position on a specific matter of statistical practice. Its message was that p-values are useful and widely misunderstood, and that no single threshold should decide what is true.