top of page
Start here
Data are statistical evidence, and they can be used to answer many different questions. One size does not fit all here.
What is it that we are measuring?
Understanding what we should measure is the first step. Selecting a tool to perform the measurement is the second. Examining the reliability of that tool is the third. Note that the reliability of the tool is not the same as the strength of the statistical evidence itself.
The standard bearer
While not as widely used as the classical p-value, the likelihood ratio (sometimes called a Bayes factor) is a flexible and reliable measure of the strength of evidence that is both foundationally and functionally sound. It takes some getting used to, however, and has never been as widely adopted as it should be.
bottom of page