Discussion questions#
The Wasserstein et al. article argues that p-values are useful in a limited manner for understanding relationships in your data. Evaluate their ATOM approach and consider whether the steps they prescribe negate the need for p-values at all.
Consider the four common poor null hypothesis test statistic practices outlined in the reading and the lecture. Each of these problems relate back to a single common misconception of what the p-value is reporting. What is this misconception?