> For example, drug use can result in being kicked off sports teams, suspensions, things that could hinder intellectual development.

Drug use can also be a symptom of a poor mental state, for example that induced by social exclusion or poor mentorship.

In such a situation the casual graph would look like:

                             /¯¯¯ Kid uses drugs
    Kid not feeling great ---
                             \___ Kid does bad in school, allowing cognitive skills to languish

I like this dialogue, because my first hypothesis (Which I didn’t want to encumber the original post with) was that , given the high base percent of users (85% or so and above), it seemed use would likely not be correlated with less social interaction, but that lack of use might be. Peer pressure doesn’t work without peers, as a DARE officer would probably say.

Maybe there are introverted individuals or closed-group social circles that are just performing better, one descriptor of those groups being the lack of substance use (Or use of a more limited set of substances, like alcohol only).

Irrespective, I think these types of questions are more useful in helping kids grow up intelligent than this study is (If we assume their focus on measuring intellect references is actually out of concern for how well students are doing)

Problem is you can't run an RCT and hand out marijuana to teens so you are stuck with biased sampling methods.

Easy personal solution is to accept the serious limitations of studies like this. Harder scientific solution is to implement a well designed cohort study that follows people over 20+ years.

That's how we know about cardiovascular health. Framingham cohort study.

I haven’t really vetted this study, but it samples IQ prior to use at 9-12 and after use at 17-20 among twins [0]. It concludes that there is no link between IQ scores and usage. More of a “For your interest” reference.

Your overall comment is spot on, just because designing a good study is hard does not mean we should accept poorly designed studies as good science. Agency risk unaddressed, science should seek to find the solution to a question rather than an answer.

[0] https://www.pnas.org/content/113/5/E500

You can do RCTs for adults, though, and teen use of cannabis is trending downward (at least in U.S. states that have legalized medical and/or recreational use).

This is why I propose we get dolphins and parrots stoned. Even if nothing is found the footage would be worth it

We tried LSD and free love already haha

https://allthatsinteresting.com/margaret-howe-lovatt

Need comparative data. Repeat same study with heavy ice cream consumption, heavy video game playing, tackle football exposure, or anything else that could impede adolescent health and development.

At first, I thought you meant repeat the study but swap out the cognitive tests so the stoned teens are playing video games and eating ice cream. Pretty sure I can extrapolate the results of that study with 99% accuracy.