Discussion of a claim (published in a top psychology journal!) on physical strength and political attitudes

The claim (by Michael Peterson, Daniel Sznycer, Aaron Sell, Leda Cosmides, and John Tooby) is pretty grandiose given the data on which it is based. The data are regression coefficients from some surveys, mostly of college students, who answer some questions about income and redistribution and have their biceps circumferences measured. The claim: “physically weak males are more reluctant than physically strong males to assert their self-interest.” To me this seems like a naive view of political attitudes, the idea that a low-SES conservative or a high-SES liberal holds these views because he is “reluctant to assert his self-interest.” Given that they’re surveying college students, I don’t know why they didn’t just directly ask about self-interest, for example views on student loan supports. But I don’t want to get into a series of criticisms here: my problem with the paper isn’t that it’s necessarily wrong but rather that it’s not at all convincing.

As a political scientist, I’m particularly sensitive to these sorts of overstated claims, which I used to associate with economists, but now I seem to be seeing it more and more in psychology journals.

7 Responses to Discussion of a claim (published in a top psychology journal!) on physical strength and political attitudes

  1. jonathan May 29, 2013 at 7:18 pm #

    Reminds me of some of my favorite work. For example, the relationship between eating hot dogs and children’s leukemia. Looks on the surface like there might be something because hot dogs are processed foods with … and as I remember the findings, the results were no connection and then a connection and then no connection, meaning that eating under say 14 hot dogs, as I remember, meant nothing and then eating more than 18 meant nothing but 14-18 was baaaaad.

    But I can’t equate this to one of the great works, the comparison on maze running of the two different AC/DC singers. If you’re not a fan, you don’t know that Bon Scott died – an actual case of choking on vomit while drunk – and was replaced by a guy who sounded like Bon Scott. I believe this was done in Western Canada – Calgary? And it showed pretty much what everyone would guess, that rats exposed to the original did better. The rats know.

    • Andrew Gelman May 29, 2013 at 8:21 pm #

      Highway to Hell, indeed.

    • Sebastian May 30, 2013 at 12:51 am #

      the AC/DC study was actually a joke and the by-product of an experimental set-up gone awry and a professor with a sense of humor.
      It was also an early example of what’s wrong with the Freakonomics people:
      http://www.freakonomics.com/2007/08/21/there-is-hope-for-economics-the-acdc-paper-was-a-joke/ (i.e. they’re ungracious & don’t care about details and context)

      • Sebastian May 30, 2013 at 1:19 am #

        oh and following up on that story, the paper actually got published in a decent journal – presumably with an editor with a sense of humor – and a great abstract:
        “These results may have important implications for settling drunken music debates”
        (and contrary to what jonathan said, the study shows that Brian Johnson is the better singer, which I think is an increasingly popular opinion. He was, after all, the vocalist on Back in Black, which is w/o doubt the greatest ACDC album).

        • jonathan May 30, 2013 at 10:49 am #

          I stand corrected. The rats were indeed “smarter” after listening to Brian Johnson.

  2. Cajunjoe May 29, 2013 at 10:08 pm #

    Similar results were obtained with respect to waist size. Conclusion: obese people like to throw their weight around.

    • Andrew Gelman May 29, 2013 at 10:19 pm #

      Believe it or not, they controlled for BMI in their regressions!