This is a reposting of a blog entry that originally appeared on 9/26/2013 at http://www.epjournal.net/blog/
Joe Henrich and colleagues’ paper, The weirdest people in the world, argued that psychology draws too much on WEIRD (Western, Educated, Industrialized, Rich, and Democratic) samples. This work has received, with excellent reason, a tremendous amount of attention.
Recently, I’ve been hearing this critique specifically about evolutionary psychology. Back in March, for instance, one prominent blogger made the following claim:
One of the common complaints about evolutionary psychology is that it claims to be addressing evolved human universals, but when you look at the data sets, they are almost always drawn from the same tiny pool of outliers, Western undergraduate students enrolled in psychology programs, and excessively extrapolated to be representative of Homo sapiens — when we’re actually a very peculiar group.
(Note that the blogger in question correctly pointed out in a comment to the post (#8) that while his post singled out evolutionary psychology, the paper by Henrich et al. did not focus specifically on the field.)
In any case, given that this is a “common complaint” – to take one more example, Kate Clancy implied that using non-WEIRD samples was one way to “make progress in evolutionary psychology” – I thought it would be worthwhile to check the claim by the blogger, namely that data sets in evolutionary psychology are “almost always” from WEIRD samples.
Before I do, it’s important to keep things in perspective. There are large numbers of social scientists who are in the business of trying to understand and explain human nature but do not take an explicitly evolutionary approach. Authors of textbooks in social psychology, for instance, when they are discussing their topic of inquiry, don’t indicate an interest in a particular subset of humanity – the undergraduates at the institutions where these scholars work – but rather they indicate that they have broader ambitions, understanding human behavior.
For instance – all bold font is my emphasis – according to Gilovich, Keltner, and Nisbett (2011), “Social psychologists go beyond folk wisdom and try to establish a scientific basis for understanding human behavior by conducting studies and setting up experiments” (pp. 7). Aronson, Wilson, and Akert (2010) define social psychology as “the scientific study of the way in which people’s thoughts, feelings, and behaviors are influenced by the real or imagined presence of other people” (pp. 3). Baumeister and Bushman (2011) pose the question, “Can social psychology help us make sense of the bizarre and baffling diversity of human behavior?” (p. 3) and answer their own question with a (“resounding”) “Yes!” Wikipedia seems to agree, indicating that “social psychology is the scientific study of how people‘s thoughts, feelings, and behaviors are influenced by the actual, imagined, or implied presence of others.” Similar remarks are made in other social psychology textbooks.
These quotations make clear that social psychologists are interested in humans, people, broadly, rather than just undergraduates at top American research institutions. This explicit focus is important because it places on equal footing the ambitions of evolutionary social scientists and non-evolutionary social scientists. Both are interested in the full panoply of human behavior, especially social behavior, both features that are universal and those that vary across populations. For this reason, social psychology provides a good comparison class. Are researchers in evolutionary psychology more or less likely to draw on non-WEIRD samples?
To address this question, I surveyed the 2012 volume of the top journal in social psychology, The Journal of Personality and Social Psychology (JPSP) and the official journal of the Human Behavior and Evolution Society, Evolution and Human Behavior (E&HB). (And by “I surveyed” of course I mean that other people did all the work. Hat tip: Fatima Aboul-Seoud and Molly Elson.) For each empirical paper, we scored whether the samples used were exclusively WEIRD, both WEIRD and non-WEIRD, or exclusively non-WEIRD. The results are in the Figure below.
Roughly 96% of the papers in JPSP, designed to “make sense of the bizarre and baffling diversity of human behavior,” came from WEIRD samples. This figure was 65% for E&HB. (A chi square on these values is hugely significant.)
Still, is two thirds enough? Should it be more? Perhaps.
Scholars collect convenience samples from undergraduates (and other WEIRDos) because they are, well, convenient. Is convenience a good reason? Maybe not, but tradeoffs are. Gathering data from the field is difficult, a fact known by no one so well as Joe Henrich, who organized the cross-cultural project in behavioral economics. Part of what made that work so valuable – conducting Dictator Games and Public Goods games, etc., in numerous small-scale societies – was that there was an array of (conveniently and cheaply gathered) data to compare it to. The Hadza Dictator Game offers (25%) wouldn’t seem so weird without the WEIRD comparison offers (~50%).
Gathering data from such small-scale societies as the Hadza requires a lot more time and effort per datum than it does among American undergraduates, making convenience samples not just convenient, but efficient. Given that there is a limited pie of social science resources – research time and granting agency funding – efficiency matters.
In some fields of psychology, gathering such data would be prohibitively expensive. My colleague Coren Apicella, who works with the Hadza, has a difficult enough time getting sufficient supplies to conduct her work (and stay alive) as it is; bringing in an fMRI machine to the shores of Lake Eyasi would be essentially impossible. Similarly, vision scientists rely nearly exclusively on samples from nearby, in no small part because of the equipment needed to for their studies. Of course, there is also the belief that their results will generalize easily to the rest of the population, but it’s worth bearing in mind that one of the examples Henrich et al. point to where results don’t generalize was in the visual domain, the Müller-Lyer illusion.
To return to the quotation with which I opened, unless two thirds of the time means the same thing as “almost always,” the (undocumented) claim by the popular blogger above is false, but the suggestion (p. 22) made by Henrich et al. seems about right:
More than other researchers in the social sciences, evolutionary researchers have led the way in performing systematic comparative work, drawing data from diverse societies. This is not because they are interested in variation per se (though some are), but because they are compelled, through some combination of their scientific drive and the enthusiasm of their critics, to test their hypotheses in diverse populations.
In sum, as measured by this analysis of samples in articles in journals, adding evolution to psychology makes the science less WEIRD, and more NORMAL (Non-Weird Organisms Resembling Mankind’s Ancestral Lifestyle, with apologies for the stretch as well as the gendered language, which I needed for the “M”). Should the social sciences be less weird still? Perhaps. I certainly hope our colleagues over at journals such as JPSP start taking the point about non-WEIRD samples as seriously as we do in the evolutionary community.
Aronson, E., Wilson, T. D., & Akert, R. M. (2010). Social psychology (7th ed.). Upper Saddle River, NJ: Pearson Education, Inc.
Baumeister, R. F., & Bushman, B. J. (2011). Social psychology and human nature (2nd ed.). Belmont, CA: Wadsworth, Cengage Learning.
Gilovich, T., Keltner, D., & Nisbett, R. E. (2011). Social psychology (2nd ed.). New York, NY: W. W. Norton & Company, Inc.
Henrich, J., Heine, S. J. & Norenzayan. A. (2010). The weirdest people in the world? Behavioral and Brain Sciences 33 (2-3):61-83.