A Tasty Correlation
Did you see the news about the relationship between chocolate and Nobel Prizes? Dr. Franz Messerli reported in a New England Journal of Medicine article that a country’s chocolate consumption is positively and statistically significantly associated with their rate of winning the Nobel Prize. If a country has high chocolate consumption, they are also likely to have many winners of the Nobel Prize.
The story was picked up by media around the world and most articles acknowledged the humor that is implicit in the situation. Many took the angle that eating chocolate makes you smarter!
Only a few pointed out that correlation is not causation. One source discusses how countries with high chocolate consumption and win the most Nobel Prizes are also those with high levels of wealth. It might be the wealth that explains this relationship between chocolate and Nobel prizes since wealth may explain both the chocolate consumption and Nobel winning patterns. Thus the relationship between chocolate and prizes is a spurious one, meaning that a third factor really explains the relationship. They are related, but indirectly – through their relationships to wealth.
Another issue that was not often mentioned in the coverage is that the data are country level data – not data from individuals. Thus a legitimate interpretation would focus only on the country level issues. To take country level data and apply it to individuals is to commit an ecological fallacy or apply data to from a macro level of analysis to the individual level. We can’t make any connections between individual behavior (eating chocolate) and Nobel Prize winning from this study.
While there may be a relationship between chocolate and Nobel prizes, unfortunately eating more chocolate will not necessarily make us become smarter! A more logical conclusion is that in countries where a lot of chocolate is consumed, more Nobel prizes will be won – but not because of the chocolate. Most likely, wealthy nations provide significant educational opportunities and also healthy funding for research. Considering the other variable, wealth, it follows that the wealthier a nation, the more chocolate eaten and the more Nobel Prizes are won.
So, how do we make use of this data? We can better understand that both Nobel prizes and chocolate are related to having resources. If a country wanted to win more Nobel prizes, they should attempt to gain more wealth – not consume more chocolate. Of course, there are cautions here since the wealth connection is not documented in these articles – it is simply mentioned as speculation.
Interestingly enough, the author of the article started plotting the data because of a review he was doing on flavanols (a component in tea, wine, and chocolate) and its positive effects on cognitive functioning. His interest in the chocolate – brain functioning relationship began with research at the individual level.
I will not commit an ecological fallacy but I will do my part by consuming some nice dark chocolate in case the relationship does hold true at the individual level. What other spurious correlations can you think of—edible or otherwise?