A study was recently published which shows that there is a direct correlation between the amount of chocolate a nation consumes, and the number of Nobel laureates from said nation. The point of the study, however, was to show how ridiculous and fallacious such studies can be.

New York physician Franz Messerli has just published a study, which shows a direct correlation between the amount of chocolate a nation consumes and the number of Nobel prizes won by people from that nation. The study's real purpose however, is not to incite an increase in chocolate consumption, but rather, to demonstrate how ridiculous a study can be when not properly analyzed

One can indeed make the jump to conclude that chocolate and Nobel prizes are related: The highest number of Nobel prizes, by nation, are Swiss, Swedish and Danish. These very same nations consume the most chocolate. The study's main concern is with validating studies by using p-values; a commonly used expression for determining the likelihood of a result based on a spread of other results. Pretty much any statistical data falls into a so called bell-curve, as seen below. The height of the curve, Y, represents the amount of statistical data found to have the result given at some point on horizontal axis X. Thus, the highest points on the curve, in the middle, represent results with a high likelihood of happening; the further away from the middle we get, the less likely.

*A bell curve. X is a particular result, and Y is the number of samples with that result*

The p-value is the probability of obtaining a result at least as unlikely as the one analyzed; in other words, the probability of any sample data getting some specific x-value. The closer the number is to 1, the closer the probability is to 100%

Messerli found that the correlation between chocolate and Nobel Prizes, though clearly "evidenced" by the numbers, is as low as 0.01%, with a p-value of 0.0001. He admits there may be some indirect correlation, such as that wealthier nations will both be able to fund more research and allow for more chocolate snacking, but that there is no evidence of an actual link between the two.

Messerli points out that research studies often go through hundreds of data sets and that it is sometimes very easy to see a connection when two unrelated items happen to exhibit signs of similarity. His lesson is simple: data is not always black and white, and though numbers can make one thing seem true, under a different light, they may be quite false. Whether this is accidental or a deliberate manipulation, one shouldn't always believe what one reads; especially if it seems too fantastical to be true.