Eating Chocolate makes you really really smart … right?

Eating lots of chocolate has been proven to make you a lot smarter.

Really?

Yep, the details are all in a study published within the New England Journal of Medicine and is entitled Chocolate Consumption, Cognitive Function, and Nobel Laureates. Read and you soon discover that the findings of the author, Dr. Franz H. Messerli, a cardiologist and director of Clinical Hypertension with St. Luke’s and Roosevelt Hospital, demonstrate that there is indeed a close significant linear correlation (r=0.791, P=0.0001) between chocolate consumption per capita and the number of Nobel Laureates per 10 million persons in a total of 23 countries.

But don’t take my word on this, here is a graph for you …

10155272_726577164030780_1826738113587635717_n

So that’s it then, case closed, because all you need to do is to start your kids out an intense diet of chocolate consumption and then sit back and watch the grades roll in.

But wait a second, this simply can’t be true?

Ah but it is … even the BBC says it is, so it must be … right?

Key Point: correlation does not mean causation.

If you are still being a tad skeptical about all this, then well done, because the “study” is of course done to point out that a correlation, however strong, does not mean causation, and so sadly no, chocolate consumption does not improve intelligence.

What is of course really going on here is that chocolate consumption, quality of universities, and Nobel prize wins are all correlated with GDP per person. A nation that invests in education will reap a benefit of churning out more nobel prize winners, but they will also probably have more wealth and so will also consume more chocolate as a nation.

You will often find claims in the media that a study has proven that X is good/bad for you and so many then rush off and start the latest fad diet in order to produce the desired result, when in fact all that the study has done is to demonstrate a correlation, and has not actually demonstrated that X actually causes whatever it is.

You can be fooled

Let’s try an even more obvious example.

Did you realise that eating Chocolate is really really … did I mention really … bad for you. The evidence is quite clear and beyond any doubt – every single person who ate any Chocolate at all between the years 1863 and 1889 is now dead, and so we have a clear correlation between chocolate consumption and a 100% mortality rate.

Hopefully you can spot the ever so subtle flaw there in that line of thinking.

Another Example

Lack of religion is associated with increased rates of depression.Therefore, lack of religion directly causes increased rates of depression.

Oh but wait a moment, people who do not believe in a god are often treated like shit and demonised by those that do, and may also be shunned by their family. It is the obnoxious behaviour that is inflicted upon them that causes the issue here, not a lack of superstitious thinking.

What is rather sad here is that many do buy into the idea that this specific correlation is a causation.

Even the experts can get this wrong

In a widely studied example, numerous epidemiological studies showed that women who were taking combined hormone replacement therapy (HRT) also had a lower-than-average incidence of coronary heart disease (CHD), leading doctors to propose that HRT was protective against CHD.

The problem was that this was in fact completely wrong. Randomized controlled trials later showed that HRT caused a small but statistically significant increase in risk of CHD.

What was going on?

Re-analysis of the data from the original epidemiological studies showed that women undertaking HRT were more likely to be from higher socio-economic groups (ABC1), and so they also had better-than-average diet and exercise regimens. The use of HRT and decreased incidence of coronary heart disease were coincident effects of a common cause (i.e. the benefits associated with a higher socioeconomic status), rather than cause and effect, as had been originally supposed.

You can find more details on this specific example here.

Bottom Line

Next time you read some story about X being good for you or bad for you (hey Daily Mail, I’m thinking of you), then step back, close your eyes, click you ruby slippers together and repeat the following three times …

Correlation does not imply causation

ice_cream_clothes

Leave a Comment