3 ms·
It's very hard to draw conclusions about causality from these sorts of a posteriori studies. That is, as you say, does cannabis cause psychosis, or do people wi
by mirimir 8y ago
It's very hard to draw conclusions about causality from these sorts of a posteriori studies. That is, as you say, does cannabis cause psychosis, or do people with psychosis self-medicate with cannabis?
In order to get at that, you'd need to identify people with and without psychosis, and then follow them for many years. See who was more likely to use cannabis. And see whether using cannabis worsened psychosis. Not an easy thing.
- shkkmo 8y agoNo, it is impossible to draw conclusions about causality from such statistics only studies. The only way to establish causality is a randomized experimental intervention. The sort of study you describe can distinguish to some extent between "A causes B" and "B causes A" but can't tell them apart from "C causes A and B"
- mirimir 8y agoI was going to make that point too, but the idea of randomly providing cannabis and placebo was too implausible.