Society advances (medically) by the use of DOUBLE-BLIND studies. This is a LONGITUDINAL study, which is of questionable value.
In a double-blind study the researchers change ONE VARIABLE (without the testers and the testees knowing) and then check how many good results vs. bad results appear.
In a longitudinal study, someone studies history. This often results in a confusion of causes. People who take vitamins probably also drink tea over coffee, exercise instead of watch TV while drinking beer, and other traits which might be the one that causes the good result.
Only look at at DOUBLE-BLIND studies that have been repeated by multiple testers. One study alone doesn’t count.
Also, watch out for longitudinal studies sponsored by indsutry groups. For example, a study that shows daily oatmeal breakfasts drops cholesterol would be suspect if done by the Oatmeal Council (if one exists). Likewise blueberry studies by the Blueberry Council, or beer studies by the Budweiser, etc.
Excellent points, Hop A Long Cassidy.
BTW, I do not always agree with articles I post, but I provide them that those who know more on a given topic than I (read: the vast majority of FreeRepublic) can winnow out the bad from the good.
I often see mistakes in causality in studies and conclusions. For example, I heard some nitwit on the radio saying that college grads have lower divorce rates, longer lifespans, etc. than non-grads. He concluded that we would get the same effects if we could just get more people into college. How about concluding that low IQs, sloth, sloppy parenting, and personality disorders correlate with every type of bad outcome?