3 ms·
Do you have any references to that Kinsey thing? I would just love to read a methodology section including that
by burnished 4y ago
Do you have any references to that Kinsey thing? I would just love to read a methodology section including that
- williamtrask 4y agoThe technique is called randomised response. Differential privacy formalises it as well
- raldi 4y agoYeah, I don’t understand why doubling the same size would compensate for instructing half the group to lie.
- mtlynch 4y agoWhat I think GP is describing is this. Say you want to survey people about something that they may feel social pressure to deny (e.g., cannibalism fantasy). You have people come in, you give them a fair coin, tell them to go into an unmonitored closet and flip the coin (only the respondent sees the result), then come back out. If they flipped heads, the question they should answer is, "Have you ever drank water?" If they flipped tails, the question they should answer is, "Have you ever fantasized about cannibalism?" If you get 55% "yes" responses, you know that about 10% of people answered yes to the cannibalism question. You know that 50% were answering "yes" no matter what, so if you double the percentage above 50%, you get the people answering "yes" to the cannibalism question.