Hacker News new | ask | show | jobs
by minikites 2840 days ago
So we shouldn't research anything related to the social sciences because some people might lie about it? What leads you to believe that smokers lie more than nonsmokers if not social science research?
1 comments

The term for this situation is ”streetlighting”, where the possible reigns over the correct.

“It’s dark in that alley. I’d never find my keys there.”

Doesn’t matter if it’s extremely difficult to get good data here. What matters is we can’t substitute bad data if that’s all we can come up with.

And although I stated it as smokers lying, the implication is that the nonsmokers are lying (by claiming to smoke).

It is my understanding that most surveys will include a margin of error. This story does not mention this, I wonder if this is just slight of hand while the east coast of the US falls into the sea?!
Or maybe the cable news sites are distracting me from this?!?