7 Reasons Everything You Know Is Probably Wrong

2. The Chocolate Study

The problem with science, is that it relies on the media, and the problem with the media, is that it has the attention span of a peanut butter sandwich.

This means that, because the success of a piece of research rests at least partially on how novel and exciting it is, there is a greater incentive to produce more and more tenuous links between seemingly unrelated concepts. If these concepts happen to play into the public imagination - whether it's cancer, obesity, superfoods, chocolate or sex - then all the better.

This is exactly what happened with the now infamous chocolate study.

This research hit the headlines, making the claim that eating chocolate can help you lose more weight compared to dieting alone. This is obviously something of a full house for the media, and was gleefully reported across the board, even in completely reputable publications. The only problem is that it was total hogwash.

The study was conducted by John Bohannon PhD (whose PhD was in the molecular biology of bacteria, not the human diet), who designed it to be purposely flawed using too few subjects, too many variables and not accounting for statistical errors. As well as monitoring weight loss, he also monitored 18 other variables including sleep quality, cholesterol and well being, so as to increase the chance that at least one of these would come up as statistically significant - it just so happened that weight loss was the one that took it. It was then published in a dodgy journal that no right thinking scientist would usually go to and released into the wild.

Although in this instance the flaws were intentional, the chances are that the next time you see a headline with the words "According to a new study...", the effect is probably the same.