Implications of teaching Bayesian statistics to undergraduate psychology students

Abstract

The current crisis in psychology is claimed to be partly caused by the way we do statistics. This is one of the reasons why it has been suggested to replace classical frequentist statistical procedures by Bayesian counterparts. However, then we must first teach it to undergraduate students. But how? Do we stress parameter estimation, hypothesis testing, or both? Do we teach Bayesian statistics using flexible programs such as JAGS, or ‘keep it simple’ with off-the-shelf methods like JASP? How do we relate it to frequentist statistics? How do we make sure our non-statistical colleagues will keep up with the students? Are they, and is academia actually ready for a Bayesian way of thinking and doing science?