If you’re a researcher, you’re familiar with the most traditional measure of significance testing: the p-value. When one assumes that the null hypothesis （帰無仮説）is true, the estimated probability of rejecting the null hypothesis is, essentially, what p-values indicate.
However, the misuse of p-values is rampant in academia, and several statistical studies have explained the challenges of using this common measure in significance testing.
To overcome these limitations and avoid common errors related to p-values, the American Statistical Association (ASA) released the “Statement on Statistical Significance and P-Values” in 2016, in which six principles on the appropriate use and interpretation of p-values were outlined.
Let us understand the ASA principles in detail, in addition to how they are applicable to research.
1. P-values can suggest how incompatible the data are with a specified statistical model