I found a lot of conclusions the article assumed to be causes of effects, were really effects of causes. For example, if a higher percentage of marijuana users have psychological disorders, maybe then people with these disorders seek to use substances, opposed to these substances inducing a psychosis.