> [I] must be a religious fundamentalist, must be an ex-alcoholic, etc.

Personally, no. Statistically, yes, at least enough for the difference to affect a study like this. That's the point of those comments.