Hacker News Reader

Why do you think that the survey wouldn't include drug dealers in the respondents?

Very few people are drug dealers, so unless you ask everybody, they will not be included.

Wouldn’t this be fixed by doing a proper, representative survey?