A simple task turns into a big adventure.

Your Bayes is not binomial enough.

Article report time! Today I read this, Five Ways to Fix Statistics, Nature (Nov 2017). It interviews six statistics luminaries about what's "wrong" with modern statistics.

Today I learned something very handy: how to encode subjective beliefs into a Bayesian prior when you only have information on the quantiles.

I've been thinking a lot about power lately. One of my favorite things!