Sunday, January 28, 2007
In a comment on my last post, the indefatigble Wisconsin librarian Dave Lull draws my attention to The Journal of Spurious Correlations: Qualitative and Quantitative Results in the Social Sciences. This seems to answer the same need expressed in the Journal of Negative Results in Biomedicine. It's all too easy to publish positive scientific results, almost impossible to publish negative ones. But negative results are at last as important as positive ones. If they are not published, then science will slow to a crawl as thousands of scientists replicate experiments that they have not been told will fail. This is a very profound issue. It demonstrates how human failings can so easily compromise the super-human aspirations of science. Or, to put it another way, it is an example of how these super-human aspirations are, in fact, illusions. Whatever we do, it will always be infected by what Christians so resonantly call 'original sin'. That said, I have always been inordinately fond of spurious correlations. Indeed, I am not entirely sure they are all spurious. Take, for example, the poached egg solution to airline safety. On any flight, you should always carry a poached egg in each pocket of your jacket. This is because no air crash victim has ever been found with a poached egg in each pocket. Call me a sentimental old fool, but I find this very suggestive.
Posted by Bryan Appleyard at 10:00 am