Quick post today.
Anyone who reads medical literature on the regular has seen “p” values. We have a basic understanding of the statistics used (I’m totally referring to myself; I have relearned statistics in a short-term form for almost every board exam I’ve ever taken).
It wasn’t until I started the deep dive into medical diagnosis and the fits and foibles thereof that I learned about Bayes theorem.
Here’s an easy link.
https://stats.stackexchange.com/q/22
The easy link looks a lot nicer than the actual equation.
https://wikimedia.org/api/rest_v1/media/math/render/svg/87c061fe1c7430a5201eef3fa50f9d00eac78810
Basically, the frequentist says “common things are common”. “When you hear hoofbeats, think horses”.
The Bayesian says, “ I hear hoofbeats. It must be horses”, but then looks out the window and says “but I see black and white stripes on those horses. Sooo, maybe those hoofbeats are zebras”. Bayesians update the hypothesis based on the new data collected.
I wonder if there are many truly excellent diagnosticians out there who are Bayesians, and don’t even know it.
I also wonder how often we miss the diagnosis in front of us because of frequentism.
Thanks for reading and considering. More to come. Peace.