Skip to content
 

Dennis the dentist, debunked?

Devah Pager points me to this article by Uri Simonsohn, which begins:

Three articles published [by Brett Pelham et al.] have shown that a disproportionate share of people choose spouses, places to live, and occupations with names similar to their own. These findings, interpreted as evidence of implicit egotism, are included in most modern social psychology textbooks and many university courses. The current article successfully replicates the original findings but shows that they are most likely caused by a combination of cohort, geographic, and ethnic confounds as well as reverse causality.

From Simonsohn’s article, here’s a handy summary of the claims and the evidence (click on it to enlarge):

simonsohn1.png

The Pelham et al. articles have come up several times on the blog, starting with this discussion and this estimate and then more recently here. I’m curious what Pelham and his collaborators think of Simonsohn’s claims.

4 Comments

  1. xi'an says:

    "implicit egotism"?! I do no see why…

  2. Lemmy Caution says:

    Nabokov said that he often got excited when he saw the word "Nobody" starting a sentence.

  3. Brett Pelham says:

    Simonsohn makes some really excellent points, but I think several of his key statistical tests are biased against the hypothesis, and many of his studies are not as powerful as they look at first blush. Nonetheless, he has certainly raised the methodological bar for ruling out confounds in archival research, and some of the statistical techniques he introduces can be used quite effectively to improve research on implicit egotism (whether to establish it more clearly or to debunk it). One has to admire what he has done.

    That being said, we are about to publish a brief rebuttal to his claims. We make two basic points. First, if you look at some of the obvious moderators of implicit egotism there is still evidence for the hypothesis, even in Simonsohn's data. For example, in the Texas marriage data, if you average together all the first names for which there is a very strong resemblance between the male and female first names studied (Eric-Erica, Carl-Carla) you see a clear matching effect even when you use the extremely strict statistics that Simonsohn prefers over our originals.

    Our second point is that the best way to rule out confounds is not to do ever more complex field studies (though that is admirable). It is to do lab experiments. We and others have done so, and such experiments yield clear evidence of implicit egotism. Simonsohn has not explained why he believes in the effect in the lab but thinks it disappears in the real world.

    We look forward to further debate about this and hope to keep the debate as constructive as possible.

    Brett Pelham

  4. A rejoinder to Pelham and Carvallo's rebuttal is also forthcoming.

    I was limited to 5-pages double space. Here is what's in them:

    Pages 1-3
    Discuss the unfortunately half-baked arguments Pelham and Carvallo raise against the "Spurious?" paper.

    Page 4
    Concrete and striking example of lack of diligence in their implicit egotism work (botched attempt to control for ethnicity in marriage decisions).

    Page 5
    Suggestion for conducting future implicit egotism work accounting for confounds in an easy catch-all way.

    Further reading:
    Rejoinder (.pdf).

    Slides for "Spurious?" paper (.pptx).

    Uri Simonsohn