This news story reminded me of some threads from a few years ago about Ron Unz, the political activist who wrote a statistics-filled article a few years ago claiming that Harvard and other Ivy League colleges discriminate against Asian-Americans and in favor of Jews in undergraduate admissions. It turned out that some of his numbers were off by factors of 2 or 4 or more, and then, amusingly, or horrifyingly, Unz offhandedly remarked that one of his more high-profile statistical claims (cited in a notorious New York Times column by David Brooks) had been based on “five minutes of cursory surname analysis.”
Unz really pulled the rug out from Brooks on that one! The world-famous Ted Talk speaker and humility expert was too proud to retract his column, so Unz had left him to twist slowly in the wind. (Hence the image above; if we’re gonna mix metaphors I might as well go whole hog.)
In all seriousness, I remain upset by the way that false claims can stick in the public discourse. Nothing new here, of course (insert Twain-attributed quote here) but particularly sad in this case in that I expect the numbers in question were not lies but rather mere mistakes.
At some point I have more to say about this case as an example of how certain intuitions can lead us astray—believe it or not, I see some connections between the reasoning of Unz and Brooks and that of authors of various unreplicated papers in social psychology—but for now, I’d just like to help out anyone who’s coming to this particular himmicane in the middle (for example, after reading the above-linked article in today’s paper) by pointing you to my previous post on the topic, from a couple years ago, which contains links to several earlier posts and some long discussions on the details of the case.