Law School Discussion

It DEFINITELY said "dietary intake" of vitamin B12 in the stimulus. You are right, if it hadn't, I would feel less comfortable with the answer.

However, as to your control group point ... you don't do a study and prescreen everyone out of the study. The chances of all the women with preexisting conditions just so happening to be in the group with lower B12 intake is really low. It is way more statistically probable that the control group (lower B12 intake) and test group (higher B12 dietary intake) had a similar incidence of preexisting heart conditions, so it doesn't affect the quality of the results. PLUS, the conclusion was that the B12 lowered the incidence of heart disease, and the correct answer addressed how it was not necessarily the B12.

Aren't you making some assumptions in there...e.g., that the risk of the preexisting condition screwing up the study is really low and that it's more probable that the groups had a similar incidence of heart disease? You are also making the implicit assumption that the test groups were large enough to allow for statistical probability of similar incidence. What if there were only 3 patients tested, 1 with the preexisting condistion and two without? The choice of the food answer necessitates these assumptions and one additional one: that the food is the source of the vitamins. The choice of the preexisting condition does not require assumptions. And I think that when you do a study, you do prescreen subjects for the very attribute you are testing the incidence of. The question is how many had it before and how many had it after? To analogize, how can you measure how far you've gone if you don't know where you started? Your test is heart disease, yes or no. That's the measure of success. Where are you putting your bad batch, or are you even considering it at all? A lack of consideration for this bias is a weakness in and of itself, with no assumptions about the nature of the source of vitamins or the composition of the study group being necessary.

OK. I think overnight I figured this one out, or maybe I forgot all the relevant details - not sure which:

42. heart disease and doctors – preexisting  heart disease condition v. nonvitamin

Here's the thing ... it was a weaken question, and the conclusion was something like, "It must have been the vitamins B12 or whatever that were associated with lower degree of heart disease."

OK ... the non-vitamin dietary question weakens this, because they are making the explicit claim not only that these women had lower incidence of heart disease, but it was because they had higher dietary intake of B12. If foods that have B12 also have some heart disease lowering thing in them, it is POSSIBLE that the conclusion is not justified, thus it is weakened.

BUT...the preexisting heart condition one is actually a strengthener ... not a weakener, because the conclusion is about a LOWERED risk of heart disease. If the women in the study had a lower incidence of heart disease (and we know they had higher intake of B12, from the stimulus) AND they had a preexisting condition, that would make the apparent effects of B12 that much more amazing, not the other way around...

Just my $.02 ... maybe I forgot a critical piece of the question when I woke up at 2AM thinking about it.

 

I do not believe you are correct. You're assuming there was no control group in the study. You can't compare one group to itself. In order to conclude that the increased B vitamin intake lowers risk, the researchers must utilize comparative data from a control group that does not have an increased level of vitamin B intake. The women in the study did not have a lower incidence of heart disease...rather, the women in the study who had greater intake of the vitamins had a lower incidence of heart disease. So I do not believe that the preexisting condition answer is actually a weakener. Any "amazing" effects of B would not be observed for every human within the scope of the research...only to the group we identify as B vitamin beneficiaries, a group that may only exist because they were the only ones in the study who did not have preexisting heart conditions.

How can you conclude on the risks of developing heart disease for a certain group if your subjects may already have heart disease before the study begins? How do you differentiate between the effects of the vitamins and the random luck of the draw of unknowingly selecting for your test or control group a patient that came into the study already with heart disease? To me, this seems like such a weakener that we don't even have to debate the issue of whether the stimulus mentioned how the vitamins were administered to the subjects (and I am sure that it didn't, a detail I double checked). I'll bet the house, farm, and accompanying silo on the already had heart disease answer choice.