Moniker mumbo jumbo

Social psychology research is known for its counterintuitive, surprising, sometimes even “cute” findings. One of the latest findings in this series is that your initials can affect how successful you are; for instance, students whose names start with C or D get worse grades than students whose names start with A or B. Authors Lief Nelson and Joseph Simmons (2007) describe this effect as a manifestation of implicit egotism, or the tendency to like things that bear some resemblance to you, in their article “Moniker maladies: When names sabotage success.” In this case, because your name is Donald and you like the letter D, you are less averse to getting a bad grade than Amy is, so you don’t try as hard. Similarly, Nelson and Simmons find that baseball players whose names start with K (the letter posted after a strikeout) strike out more than other players, students with C and D names attend lower rank law schools than A and B name students, and people whose initials match a consolation prize solve fewer puzzles.
Sound too good to be true? Unfortunately, it is. A scathing analysis of Nelson and Simmons’ results (McCullough & McWilliams, in prep.) reveals that multiple misapplications of statistics and a few just plain odd assumptions actually account for the results. Take the baseball letter K finding, for example. Nelson and Simmons compared the strikeout rate of players whose names start with K against the average of all other letters. When the analysis was re-run using other letters, McCullough and McWilliams discovered that all initials except C, M, R, U, and V were statistically significant. Basically, any letter you test is likely to correlate with more strikeouts than average or fewer strikeouts than average simply because of variance from the mean. But, of course, the original authors didn’t test the letters that weren’t convenient to them.

Or the GPA example. The natural hypothesis would be that GPA(A) > GPA(B) > GPA(C) > GPA(D) > GPA(F), but that’s not what Nelson and Simmons tested. Rather, they combined A and B into one group, C and D into another group, and left out F altogether. According to a footnote, they “did not consider F initials to be grade relevant because, compared with A through D, F is much less universally associated with an academic-performance outcome” (p. 1107). I’m not sure how that one got past reviewers, but it seems like if you’re looking for initials associated with grades, F should probably be one of the first letters you try.

And the list goes on. The point here is not that implicit egotism as a whole doesn’t exist; there is a wealth of strong evidence that it does. However, the way in which it has been misapplied in this article is troubling. Do reviewers really sacrifice rigorous examination of the methods and analysis used for a cute and memorable set of conclusions? Let’s hope “Moniker maladies” was just a fluke.