The following relates to a study of teenagers who’d been taken to a mental hospital and told they were mentally ill, and then compares the frequency of recorded physical altercations between those who had previously smoked pot and those who had not. In this group, those who had previously smoked pot were more likely to have a non-zero frequency of a documented physical altercation in the next 12 months.
The authors conclude that cannabis causes violence. (I’m not linking to the study.)
a) If you’re going to think I’m crazy anyways, then why not act crazy? This could be habit forming in a sense and explain the results. If you’ve ever been a teenager or met a teenager, you may be able to understand that this is a pretty reasonable claim. Sure, you can try to patronize me or humiliate me in numerous ways to try to force the result of abandoning “crazy” behaviour, but in teenagers this could easily cause lashing out, depression, or other such things.
b) The fact of being illegal puts much of the trade into the hands of organized crime, which then increases the probability of trade-related violence.
c) If you i) believe the study result at face value and ii) apply it in biased manner in real world decisions (e.g., to assume that the “causal” relationship would apply to every single case, as opposed to, say, the 1% of cases identified as different), then your own natural mental defects related to processing bias, etc., will at the very least cause the result to be exaggerated, because the individual who has previously smoked cannabis is more likely to get a record for the same act that another person wouldn’t.
I personally believe that cannabis causally operates in the other direction (less, not more, violence prone). All of the above explanations show how that can be true even if the statistical results are factually correct with respect to i) considering the subset who’d received a diagnostic opinion of mental illness followed by ii) the differential of recorded incidents of physical altercation between those who’d smoked pot before and not, in particular due to the effects of b) and c).
If you cannot acknowledge the direction of effect, if not necessarily the relative magnitude, of b) and c) — not necessarily in a paper, although I think you should, but I mean internally, in your head, understand that this is OBVIOUSLY the case — then you (researchers) shouldn’t be within 1000 miles of telling people who’s defective and/or why.