3
$\begingroup$

I came across this post http://xkcd.com/1132/ and this image

enter image description here

I didn't get why the Bayesian statistician is saying no. Any suggestion

$\endgroup$
5
$\begingroup$

The idea is that the Bayesian has prior information suggesting that the probability that the machine is lying (1/36) is much higher than the probability that the sun exploded (about 0). So he assumes that the machine is lying. The frequentist, meanwhile, notes that the observed result would have been unlikely under the null hypothesis that the sun were still intact (the machine only lies and says "yes" about a 3% of the time in that scenario), so he rejects this hypothesis.

Andrew Gelman, a Bayesian, doesn't particularly like the comic, though. There's some good discussion in the comments of the page I linked to (including a comment by Randall Munroe, the author of XKCD).

$\endgroup$

Not the answer you're looking for? Browse other questions tagged or ask your own question.