On: Dramatic drop in marijuana use among U.S. youth over a decade

[elliottkember]

> the skew is likely to stay the same over decades

This is a huge assumption.

[hx8 → elliottkember]

Yeah, I don't agree with that at all. A simple change such as "More schools are giving the survey right before standardized tests instead of at the end of the school day" may skew numbers over years.

Edit: There's probably dozens of variables with how the survey is given that will skew the bias. They probably don't even ask the exact same question in the exact same way on the physical survey.

[mikem170 → hx8]

> A simple change ... may skew numbers over years

Is this your guess, or is there other conflicting data?

[hx8 → mikem170]

Am I the only one that questions the validity of self reported survey data? No, the concerns are extremely well documented, including documentation about psychological and situational factors that impact the results.

[mikem170 → hx8]

So it's possible that the numbers have changed even more than this study shows?

[hx8 → mikem170]

Sure, I'm not arguing in which direction the numbers reflect Truth. I'm simply saying I trust teenage self reported cannabis usage about as much as I trust self reported exercise habits. I'm very skeptical, and I don't think we can assume constant variance in self reporting across different time and institutions.

[mikem170 → hx8]

Does that mean all self reported studies are worthless, in your opinion?

[elliottkember → mikem170]

The point is that we don’t know, and this means the data may be wrong. For an experiment to prove something, you need to control for these variables.

[moomin → elliottkember]

You can’t do experiments on entire populations. That’s precisely why we have studies in the first place.