🔤DEFINITION AND ORIGIN
The probability of getting data at least as extreme as what I observed, if the null hypothesis and the model's other assumptions were true. A small p says the data sit awkwardly with that assumption.
From 🎓 Lecture 8 - Hypothesis tests and what p means, the week my midterm mark came back. 🔬 The ASA Statement on p-Values is the document to reread before an exam.
Origin: p for probability. The habit of treating 0.05 as a convenient line goes back to 🏛️ Ronald Fisher in the 1920s, who offered it as a rule of thumb.
Where it leads: 💎 A significant result is not the same as a true one. I drilled this with the Statistics tutor prompt until I stopped saying the first wrong thing below.
Three things it is not
- Not the probability that the null hypothesis is true.
- Not the probability that the result came about by chance alone.
- Not a measure of how large or how important the effect is. A tiny effect in a huge sample gets a tiny p.