3 ms·
Bayesian Adjustment Does Not Defeat Existential Risk Charity
- gwern 12y agoI submit this not because I think HN really cares all that much about existential risk charity, but because this illustrates an interesting way that statistical arguments can be flawed: we're generally fairly savvy about how p-values go wrong, the need for power, and so on, but it's much rarer to see someone object to the choice of distribution. In this case, the choice of a normal distribution (or any thin-tailed distribution) almost entirely determines the result of the original argument, and inasmuch as there is no clear justification for why we would make the assumption that charity efforts follow a normal distribution, the original is just a very long exercise in begging the question.