Is Evolutionary Psychology WEIRD or NORMAL?

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.

Fig. 1. WEIRDness in two journals.

Fig. 1. WEIRDness in two journals.

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.

25. September 2013 by kurzbanepblog
Categories: Blog | 17 comments

Comments (17)

  1. Apologizing for “gendered language” is so very WEIRD of you.

    Given the history of the word it’s not actually “gendered”

    Or maybe I’m just a NORMAL patriarchal oppressor.

  2. There’s a possibly-underappreciated irony here. Recently, the following metaphor was put to me:

    “Would you ever conduct a study with an N of 1? Probably not. Why? Simple: N of 1 is probably in a poor position to generalize to the population.”

    While that metaphor was applied towards a specific research topic of mine, it applies here as well. This “unnamed blogger(s)” who made the claim is drawing a conclusion about a field from an N of 1. Evolutionary psychologists tend to use WEIRD samples. Well yes, but compared to what? That conclusion is, well, not very cross-cultural, to the extent that psychology contains different research cultures.

  3. I think E&HB is not the best choice for this comparison. It is a multidisciplinary journal where plenty of anthropologists publish and my sense is that many of them would disagree with their work being labeled “evolutionary psychology.” I think “Evolutionary Psychology” would be a more applicable comparison journal.

    • I initially had a thought similar to Matt’s. Then I considered how several researchers with Ph.D.’s in anthropology (e.g., Carlos Navarrete, Coren Apicella, Mike Gurven, Joe Henrich, Larry Sugiyami – I think that I’m accurately representing the background of these individuals, though I’m not certain), many of whom work in anthropology departments, have done and continue to do work that (I think) most would classify as “psychology” (e.g., Gurven’s work on personality among the Tsimane which, funny enough, was published in JPSP rather than EHB last year – note that my point is not that these researcher only do psychology work, but that some of their work does seem to be “psychology”).

      I suppose you’d have to survey individual researchers with anthropological backgrounds or affiliations, but I’d guess that many would have no objection to (some of) their work being labeled as “psychology.” Further, if they’re publishing in EHB, they’d probably label this “psychology” as “evolutionary”. Naturally, whether they’d agree with personally being labeled “evolutionary psychologists” is another issue.

      • As an example in the latest issue of EHB there is an article by Adrian Bell and Tim Waring with experiments in India. Adrian and Tim have degrees in ecology. Adrian is in an anthropology department (Utah) and Tim is in an economics department (Maine). The experiments in their paper are more clearly in the experimental economics, not the psychology, style. Classifying this work or their research generally as “psychology” would be a stretch and they probably would not self-identify as EP researchers. Should this work be counted as EP research?

        What if we limited the above analysis to self-identified EP researchers? My guess is that the tested sample would be more in-line with social psych.

        Generally, I find the argument that “we are doing it better than the social psychologists” weak sauce. I guess if the goal of science is to do *relatively* better than scientists in other sub-disciplines this would have merit. But I take a more absolutist view of the goals of science.

        • Interesting points, Matt. I suppose that whether one interprets these data as “weak sauce” substantially depends on how one defines and categorizes “evolutionary psychology.” I take Rob’s last comment above – “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” – as a bit more inclusive than your interpretation. I think one of the (unstated in this blog post) strengths of evolutionary psychology (or evolutionary behavioral sciences in general) is the “fuzziness” of traditional disciplinary boundaries, since there is significant (though not perfect) overlap in the theoretical assumptions underyling evolutionary approaches to behavioral economics, social psychology, personality psychology, developmental psychology, behavioral ecology, demography, etc. The methodological traditions in these different fields can be applied to testing the same underlying theoretical models.

          Again, acknowledging the subjectivity in this issue, I’d suggest that a researcher’s use of methods developed by experimental economists does not preclude their work from being categorized as evolutionary psychology, nor does the department in which an individual works. This paper immediately pops to mind, though there are myriad other similar examples:

          Kurzban, R., DeScioli, P., & O’Brien, E. (2007). Audience effects on moralistic punishment. Evolution and Human Behavior, 28(2), 75-84.

          • Josh,

            These are all fair points. I am not normally one for disciplinary distinctions, but the above post is *all about* disciplinary distinctions. It divides the literature into two disciplinary categories and counts the number of non-WEIRD studies in each. I agree that such divisions are to some extent arbitrary, but that is my point. The above results depend on how you define EP, and defining EP as including its chief critics does not seem like the most obvious way to do it. This seems strange to me:

            Evolutionary Anthropologist: Evolutionary Psychologists should use more non-WEIRD samples, like we do.

            Evolutionary Psychologist: That’s easy, we’ll just define all your work as Evolutionary Psychology. Then we will be, on average, better than the social psychologists.

            Evolutionary Anthropologist: Alright then, carry on.

            The part I find weak sauce is the argument that as long as EP is better than social psychology, on average, then everything is ok.

        • ” I find the argument that “we are doing it better than the social psychologists” weak sauce.”

          I don’t mean to speak for Rob here, but I think this is not the point at all. If EP is *singled out* as awful because of X, then it’s reasonable to ask, is X a property of EP, or just a property of social science? If it’s the latter, the criticism is invalid and disingenuous no matter the content of X. I also don’t think the social psychologists are doing it “wrong”. They might be somewhat careless with language. If the typical SP has the perspective, “my finding applies to all cultures on earth past and present” then that is a problem. But I don’t think you’d hear that.

          Also, as it happens, the criticism is perversely mistaken. EP, more than any field has universality criteria in its DNA, critical to findings being considered robust.

          I see no reason to get hung up on the “P”. My BS is in psychology, I’m now part of the bio section of my anthropology department. I am an evolutionary psychologist and an anthropologist, and so what. For purposes of this discussion, the critical part is the evolutionary perspective: does your hypothesis have as a critical component natural selection? That’s about it. I don’t really care if you’re a neurologist or economist, you can still speak to the topic. No one has yet kicked either out of topics like philosophy or politics.
          And one other thing… like Obama once said, he calls himself black (not “mixed” though he has a white parent) because he is perceived as black. If we’re responding to critics who take anyone using Darwin to explain behavior to be “EP”, then it is our right to reply using those same terms (or to depose them as foolish if they’ve simply become very confused about who is saying what and why).

  4. Rob, this is awesome.

  5. Pingback: Post Of The Week – Thursday 26th September | DHSG Psychology Research Digest

  6. Nice post. I have a feeling that the beginning salvo that set the evolutionary cat among the psychology pigeons was in fact issued in part by an anthropologist. I know this might annoy some people in EP. but it would do many of you a lot of good, and save a lot of time, if you simply availed yourselves of some of the ethnographic studies out there. And these include lots of studies of people in WEIRD societies…

    Aside: Unfortunately, there was some degree of disgruntlement with a particular school of cultural anthropology in that salvo, that got packed into a straw man (the social science socially constructed behaviour model or whatever it is called) to add flair. I had never heard of this before I was told this approach existed by someone in EP. In over 40 years as an anthropologist, it never once appeared to me that there was anything BUT an evolutionary paradigm in my discipline.

  7. To reply to some of the comments here, to clarify, I did not argue that evolutionary psychology is “doing it better” (broadly) than social psychology. I used social psychology to get at the issue of whether having an evolutionary perspective increases or decreases the use of non-WEIRD samples. As I said in the post, the agenda of social psychology makes it a good comparison class, and social psychology (JPSP) provides one benchmark. Because there is no objective standard for how much work ought to be non-WEIRD, as far as I can tell, social psychology is useful to compare against. As for the issue of classification, for this particular purpose, to me the issue is less what one calls oneself; rather, the issue is whether a scholar uses evolutionary ideas to inform their work. Finally, I felt E&HB and JPSP were reasonable choices to look at because of their prominence. In any case, thanks for the comments, all.

  8. “to take one more example, Kate Clancy implied”.

    This link is screwed up.

  9. I cannot resist pointing out that one of those 3% of 2012 JPSP papers that used a mixture of WEIRD and non-WEIRD samples was my paper on the “Binds and Bounds of Communion”, and the key idea guiding my hypotheses (that “communion has bounds because the binds of communion have costs”) was explicitly derived from Kurzban & Leary (2001). More generally, despite my primarily doing “standard social science”, I am indebted to evolutionary psychology for many things, including keeping the question of universality salient in my thinking.

  10. Pingback: Regrettable Coverage | Evolutionary Psychology

  11. Pingback: Why you shouldn’t trust 96% of Evolutionary Psychology | EvoAnth

Skip to toolbar