<link rel="stylesheet" id="wp-block-library-css" href="https://c0.wp.com/c/5.8.1/wp-includes/css/dist/block-library/style.min.css" type="text/css" media="all">
loader image

Evolutionary psychology (EP) attempts to understand the evolution of our psychology. No surprises there. It’s an important field that could shed a lot of light on both ourselves and our ancestors. Unfortunately it’s a field rife with issues; primarily it’s use of unrepresentative samples. It’s hard to draw conclusions about all of humans (and our evolution) when you only study Westerners.

Over the past few weeks I’ve been exploring this problem. In part 1 I revealed the vast majority of EP research relied on Western (or WEIRD) samples. In part 2 I discussed whether or not evolutionary psychologists had addressed this problem in recent years (answer: they’ve talked about it a lot, yet the use of WEIRD samples has not declined).

Today is the third and final installment: why all of this matters.

The problem

Sure, evolutionary psychology uses a lot of WEIRD samples. But does this matter? I mean, Westerners aren’t representative of humanity as a whole (being more easily fooled by illusions, for example) but are they different enough to invalidate EP research that examines them? Have I discovered a real problem in my first two posts, or am I making a mountain out of a mole hill?

The Muller-Lyer illusion. The lines are the same length; but people from different cultures have differing levels of success at figuring out this is actually the case

That’s the issue I wanted to examine in the final part of my research. If I do research on WEIRD people, can I generalise those results to all humans or not? One easy way of examining this would be simply to look at how often research on non-WEIRD people contradicts research on WEIRD people.

Unfortunately this is not a perfect way of testing this, given how complex human psychology is. Sure, Westerners may respond to optical illusions differently, but you can’t really use to that to predict whether their fear of snakes is universal (and so maybe evolved) or not.

There’s also issues surrounding publication bias. Research which simply repeats a previous study is a lot less likely to get published than new research. Similarly, research with negative results is also less likely to see the light of day. These create the “file-drawer effect“, which can make it very hard to reliably say how often WEIRD results are challenged by non-WEIRD ones.

In short, take the results presented below with a pinch of salt. I still think they’re important and worth discussing, but it’s unlikely this is the final word on the subject.

The aforementioned salty results

Hopefully by now you know the drill: I examined research published in several EP journals from 2003 – 2007 (and in 2013), examining how much of it used WEIRD samples and how much didn’t. When looking at the non-WEIRD research, I then examined whether or not it challenged or vindicated research previously done on WEIRD samples.

The result? WEIRD results were actually vindicated by non-WEIRD research in the majority of cases….by a whisker.

evolutionary psychology

The aforementioned results, in handy graph form

In total, 56% of non-WEIRD research supported WEIRD conclusions. Given ~75% of EP research is done on WEIRD samples, if you extrapolated this data (which, as I mentioned earlier, is problematic) it may well be that 1/3 of all EP research is just wrong. Invalidated by the use of unrepresentative samples.

In short, the issues surrounding WEIRD research is not simply (for want of a better word) academic. It represents a real, significant problem that could be undermining a large swathe of evolutionary psychology. Despite that there’s no evidence the field is fixing itself. Examining data over the past few years reveals no significant change in the use of unrepresentative samples.

No positive spin or joke to close out on. Evolutionary psychology has a major problem. It isn’t getting better.

It needs to.

Related posts


Chris Reynolds · 27th January 2015 at 9:34 pm

Before your posts I hadn’t come across the term “Evolutionary Psychology” before and have read your posts with interest.

Would I be correct in assuming that if tests are carried out on different groups of individuals this shows that the psychological behaviour is due to cultural differences? You are pointing out that if tests are confined to people of one cultural background you can’t assume that all people would behave in the same g. I fully approve of your criticism – which would apply to any field of science requiring proper random sampling.

But of course if they find all humans behave in the same way it may mean no more than that all humans share the same area of culture and have nothing to do with genetic evolution. My own research – modelling ways in which information held at the neuron level might be translated into intelligent thoughts – suggests that culture is far more important than genetics.

So if all humans show the same results in a particular psychological test how do the Evolutionary Psychologists decide if the result is due to culture or genetics?

    Adam Benton · 14th March 2015 at 2:49 pm

    Identifying that a behaviour was present cross-culturally wouldn’t prove it was evolved, but it would be a solid step in the right direction. A step too little research takes at the moment.

thomscottphillips · 27th January 2015 at 10:16 pm

Please stop.

I was one of the commentators who commented on your first post, pointing out that EP is better than Psychology as a whole when it comes to cross-cultural comparisons. After all, Psychology as a whole has the same goals as EP: to understand the architecture of the human mind. And when Kurzban made his comparisons, he found that EP is doing much better when it comes to using non-WIERD samples.

Your analyses seem to confirm this: you wrote in a previous post that “a statistical comparison with the results for the entirety of psychology… reveals that evolutionary psychology uses Westerners significantly less often… EP research utilises university students about 40% of the time, which is half the rate (~80%) of psychology as a whole”.

In other words, EP is doing much better on this than Psychology as a whole is. EP is not perfect, but the idea that WIERD samples is a particularly EP issue is just false. So why aren’t you writing that Psychology has a problem? Why single out for criticism one area which is in fact ahead of the curve?

Science is about comparisons. If we are told that something occurs 40% of the time, we have no real way to know whether that is lot, or a little, without some comparison measure. In this case, when we get that comparison measure, it turns out that 40% is a lot less (i.e. a lot better) than the norm. Yet your headline announces practically the opposite conclusion. Reporting the opposite of what your data actually say is not science. It is close to trolling.

    jonyantunes · 26th May 2015 at 5:38 pm

    In the case of EP it becomes increasingly important to look for non-WEIRD samples than in general psychology, as you preety much are trying to infere behaviors and cognitions as representative of human nature.
    In fact, tt’s even more important than that: With guys like Gaad Saad pompously annoucing that “All psychology is evolutionary psychology”, meaning that our behavior can be on it’s majority traced back to an evolutionary explanation one way or the other, it means that evolutionary psychology is a foundational field of psychology. It Is the theoretical root of general psychology. If evo psych wants to take credit for that then it becomes increasinglly needed to be has rigorous as possible in their methodologies, meaning +Non-WEIRD samples. With great power (or, in this case, credit) comes great responsability.

    One should indeed focus WAAAAAAAAAAAAAAYYYYYYYY more on Non-WEIRD samples (let’s also not forget that non-WEIRD samples account for the overwhelming majority of the world population). In fact, non-WEIRD samples are not even enough. One should study the remaining hunter-gatherers tribes that have been untouched by modern culture as much as possible. It’s following the ethological principle of studying animals (or humans in this case) in their natural habitat. WEIRD samples should only be used.
    WEIRD samples should therefore be used to confirm results and hypothesis found in hunter-gathering tribes.
    Studying genes, like in behaviroal genetics, should be a common practice for infering that an emotion, behavior, etc is indeed a product of evolution and not culturally determined.

    In fact, samples and methodologies are not the only problem with evopsych.
    It ranges from ignoring paleoanthropological evidence, to come up with unfalsifiable post-hoc arguments, to assuming that every behavior is an adaptation, to assuming that correlation implies causation, and much more…

    EP is trolling in many way. It is weak as a science althoug i think we till better off with it than without it. The way it is nowadays it’s kind of a clusterfuck really…

Norman · 5th February 2015 at 8:08 pm

How about this evopsych explanation of our preference for mowed lawns?


    Adam Benton · 14th March 2015 at 2:46 pm

    Well, it’s on the internet so it must be true

Leave your filthy monkey comments here.

This site uses Akismet to reduce spam. Learn how your comment data is processed.