Originally posted by dadudemon
In in depth study used to discoer rare reactions to smoking MJ would certainly take lots of time and money. The sample size would have to be in the thousands and include as many races as possible.
It is more complex than that even. The race thing is good, I wouldn't have even considered that, but like, people's medical histories (not just ailments, but the medication they took), marijuana use, other drug use, etc would have to be controlled for. I think gender would be a very important one to look at, as I know few girls who get high the way boys do.
Smaller sample sizes might be ok, since it is more important to have replication than a single definitive study. It all depends on the "probability of correctness" [p value] you want, and if it is a strong effect, you can probably get a significant one with maybe 2 dozen people. Large sample sizes also suffer from false positive problems. Large samples can make non-significant effects appear significant, though this is best if you are looking for that 1-in-1 000 000 marijuana effect.
Originally posted by dadudemon
Also, in order to REALLY do any good research, we would have to follow MJ use for many years. Decades? Maybe.
the problem is more one of ethics with this. You could take a child, make them smoke pot every day, and within a few years you would have a good result. Not a very long study, especially considering how long medical studies usually take.
However, the current ethical regime in science sort of forbids this sort of thing. I've never tried to run a study like that, so I can't say where the gray area is or whatever, but the harm caused by the drug makes it unethical to subject participants to it. Further, because of this, the only people who can be studied as "pot smokers" are people who are pot smokers by choice, and not by random assignment.
Again, not uncommon, as there are many things that can't be randomly assigned, like gender or age, yet these are studied in controlled settings all the time. Why drugs are so difficult is because, and I'd go on a limb to say at least 30-40% (pot included) of drug use is symptomatic of other physical, mental or social problems in a person's life. So this runs into the first part, about how difficult it would be to find people who are just "pot smokers", especially given the ethical concerns with randomly assigning people to a "pot smoker" experimental group.