Hacker News new | past | comments | ask | show | jobs | submit login
Statistics Done Wrong (refsmmat.com)
2 points by felipellrocha on Oct 21, 2012 | hide | past | favorite | 1 comment



Though the points made in this article are true, it is basically making the "frequentist vs. bayesianism" argument without mentioning Bayes' Theorem.

In other words, using the included XKCD picture as an example, the likelihood before any experiments were done that "green jelly beans, specifically, cause acne" is very very very small, because it's just a silly idea without an obvious causal mechanism. Therefore, even after the experiment is done, this likelihood is still vanishingly small (though larger than before.) The fact that it had "power" as per the p test is a red herring.




Join us for AI Startup School this June 16-17 in San Francisco!

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: