Yeah, I just can't imagine the diet having a strong influence. I hate these studies that "show a link". I can do a study that "shows a link" between playing badmitton as a child and having strokes at the age of 90. How MUCH of a link? How MUCH of an effect? I discount any study that isn't willing to answer those two questions and that hasn't been able to independently verify the results.

Many of these "studies" are done by looking at a single large sample of people. No matter how large the sample, there are going to be variations in every parameter and purely coincidental correlations in the data. So what you do is you duly note the apparent correlations and then you take another, independent, sample of people and see if those same correlations hold.

When I was taking prob/stats, our prof told us about a classis misuse of correlations. It seems that some grad student noticed a correlation (back in the 60's or 70's, I think), and a pretty strong one, between a student's performance on the ACT and the altitude of the testing center where they took it. They took several samples and they all showed a similar correlation. The higher the testing center, the worse the performance. Soon, all kinds of "experts" were on the bandwagon about how the lower oxygen levels at even modest elevations having a very noticeable impact on intelligence. Finally, another grad student somewhere else thought to test this hypothesis by seeing if the effect held for other measures (which, apparently, no one else had bothered to do before becoming a "talking head"). So he repeated the analysis but looked at the SAT - and sure enough, there was a pretty strong correlation - in the OTHER direction! Students at high altitude did BETTER on the SAT. So now he looked at lots of other indicators (the Iowa Basic Skills test and a host of others) and found that the correlations were in the noise.

So then he looked for other correlations involving the SAT and ACT and what he found was that older, more established schools tended to rely on the ACT while relatively newer schools tended to favor the SAT - for a host of reasons most of them having to do with tradition. Well, guess what? This country was settled from the coasts inward and the universities followed that trend as well. The school's at the lower elevations, out of pure coincidence, tended to favor the ACT and the school's in the higher interior of the country favored the SAT. Since most students that go to college do so at local colleges, the highschools inevitably (and usuallyunconsiously) skew their teaching practices toward the test that is more likely to be important to their students.

Even as late as the early 1980's, I remember that this correlation existed. Coming from Colorado, we had practice exams and study sessions and almost all of them focussed on the SAT. I also remember that MIT and other schools that I was applying to back east only accepted the ACT while schools here such as the Academy and the Colorado School of Mines accepted both but always talked in terms of the SAT. I think this distinction has largely dissappeared now and most schools accept either without much of a preference.