Hacker Newsnew | past | comments | ask | show | jobs | submitlogin

Negatory - the sets being compared aren't independent.

That is, the overall population is (roughly) half red and half blue. If you divide that population into groups - on whatever basis - and the largest group is 93% red, then it follows that some of the other groups are going to be mostly blue.



That's only the case if you have equivalent populations and people can make only one choice, neither of which is true here.


They're equivalent to a first approximation, and the numbers we're talking about are from a poll where respondents could make only one choice.




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

Search: