Does chocolate cause Nobel prizes?

This has been around for a couple of months, but I just discovered it. An article in the New England Journal of Medicine notes a correlation between per capita chocolate consumption and per capita Nobel prizes in various countries:

The article’s clearly meant to be playful, not serious — something I didn’t now NEJM went in for.

This result came to my attention via a report on the BBC program (or rather programme) More or Less, which includes the following:

Eric Cornell, who won the Nobel Prize in Physics in 2001, told Reuters: “I attribute essentially all my success to the very large amount of chocolate that I consume. Personally I feel that milk chocolate makes you stupid… dark chocolate is the way to go. It’s one thing if you want a medicine or chemistry Nobel Prize but if you want a physics Nobel Prize it pretty much has got to be dark chocolate.”

But when More or Less contacted him to elaborate on this comment, he changed his tune.

“I deeply regret the rash remarks I made to the media. We scientists should strive to maintain objective neutrality and refrain from declaring our affiliation either with milk chocolate or with dark chocolate,” he said.

“Now I ask that the media kindly respect my family’s privacy in this difficult time.”

The program goes on to talk about the actual lesson here, which is our old friend Correlation Is Not Causation. This takes me back to the glorious moment when I first discovered the existence of xkcd:

 

In the case of the chocolate-Nobel connection, the correlation is highly statistically significant — that is, it’s overwhelmingly unlikely to get such a correlation by chance. That could be explained if chocolate consumption caused Nobel prizes (or vice versa), but it could also be explained if one or more other factors cause increases in both. More or Less explains this pretty well, but oddly doesn’t mention what seems to me to be the most obvious such factor.

Chocolate, despite what you may hear in some quarters, is not a biological necessity but rather a luxury. So you’d expect chocolate consumption to be positively correlated with wealth. And wealthy countries have resources to spend on science, so you’d definitely expect Nobel prize rates to be positively correlated with wealth. So the link between Nobel prizes and chocolate may simply be an artifact of the link between each of these and wealth. I’d bet that that’s all that’s going on here.

4 Responses to “Does chocolate cause Nobel prizes?”

  1. It does look like there is a logical connection between chocolate and Nobel prizes, but what would happen if we corrected the supplied p-value for multiplicity? Perhaps the author of the graphic looked at a million factors before finding this one.

    Looking at the graph, I wonder what colour chocolate they prefer in Germany.

  2. Brent Follin says:

    How statistically significant is the correlation though? If we’re including chocolate in the things to compare to Nobel Prize winners, one would imagine the number of possible correlations we care about is rather large. I’ve always assumed the look-elsewhere-effect as the explanation for this plot.

  3. Ted Bunn says:

    Often, this is the right explanation for this sort of thing, but it doesn’t look that way to me. The formal p-value is under 10^-4, so you’d have to have sifted through a lot of candidate correlations to find one this strong. I don’t see any evidence that the authors did this, and it’d be an awful lot of trouble to go through just to get a quirky but obviously unimportant result.

    They did have a motivation for checking for this particular correlation: there is “real” research connecting some compounds in chocolate with improved cognitive function in animals. Looking at the correlation with Nobel prizes based on this is obviously whimsical, but it’s not a totally random choice.

  4. The obvious outliers are Sweden and Germany. As a fraction of GNP, Sweden spends more on basic research than do most countries of similar GNP per capita, and Germany spends less. (Germany can probably “get away with this” in some sense since it is, by population, the largest country in Europe and thus does well in term of absolute numbers (which is what most people care about) even if the per-capita ones aren’t that good.) Another explanation is that research in Germany is organized at the state level, rather than at the national level, as a result of exaggerated federalism, and this shows why that is wrong. Sweden, in contrast, is very centralist.

Leave a Reply