That survey is with the same group of people over a time span of years...seems like a very odd way to sample things.
Not really. If they're interested in an increase or decrease, then a paired test makes sense. If it's not paired you might see that someone's smoking 3 cigarettes a day at 29 years old and it seems small, but maybe when they were 18 they were only smoking 1 a day. You have no reference. They can also use this same type of study to find out which types of people change or forget their original answers so they can improve how they do surveys. I'm guessing part of the reason they do it this way is because people who answer a question at age 29 about their smoking habits at age 18 are likely to remember things wrong. So it's probably better to do it this way.
well, for every study conducted on people, there will always exist at least one way to dismiss the findings whether correct or semi-correct. this here won't be different.
but that doesn't defeat the general premise. the fact that the initial apparent evidence based on what appears in front of us daily will point to one thing does not mean that that thing will be true. more often than not, it isn't. especially when it comes to stereotypes and race/ethnicity. specifically for weed and other drugs, it shouldn't be a hidden fact that the highest percentage of end-users of weed and other drugs is not black people. it just so happens that black people get in trouble far more for it than others.
lol i'd say this thread has officially been derailed.