Does psychology's over-reliance on American undergraduates distort our image of the human species?
Imagine that you’re in a room with 100 psychopaths. The first thing you’ll probably want to do is leave that room. However, once you do, you discover a booth installed with one-way glass where you can watch what’s taking place without anyone seeing you. Comfortably seated, you observe a strange experiment taking place. A few of the individuals have on white coats and are carrying around clipboards while most are being run through a battery of psychological tests.
Slowly the frantic activity begins to make sense. Some test subjects are looking at video monitors and have sensors attached that measure their galvanic skin response to the images they see. Others are being given questionnaires to elicit their answers to a variety of social situations. Still others are being placed inside an fMRI scanner to measure the blood flow in different regions of their brains. All of these are standard methods in the psychological and brain sciences. But what’s most striking to you is the fact that this study is being conducted on psychopaths by psychopaths.
"Subjects reported a consistent disregard for the feelings of others and a lack of remorse in cases where they've hurt someone," reported one researcher from his report based on answers from the questionnaire.
"This is consistent with the fMRI results that show significantly less blood flow to the paralimbic system, especially those regions involving emotion," adds another looking at her analysis of the brain scan data.
"The skin conductance data also agree, showing little or no emotional reaction to violent or disturbing imagery," reports a third who seems to be the one in charge of this strange experiment.
"These results suggest that the human species is inherently deceitful, antisocial, and has little regard for others," he says. "Evolution has honed us to be selfish actors interested only in maximizing our individual potential at the expense of everyone else." The other researchers nod in agreement, for that is certainly what the results show.
From where you sit it's clear that something is terribly wrong with this study. Because they were only testing psychopaths the researchers' data may be consistent but it's only applicable for that one group. However, because the researchers were also part of that group and saw the world in the same way, they made the false assumption that humans everywhere behaved that way too. This is known in the sciences as confirmation bias, preferring conclusions that support someone's own personal preferences or outlook even when the evidence is weak to nonexistent. This usually happens unconsciously. It's the tendency we all have to prefer interpretations that support our preexisting beliefs. This is why scientific studies try to get a large and diverse sample size to draw their conclusions from.
Obviously the above example could never happen in real life, but it represents a simplified thought experiment to address a larger question about how research on human cognitive evolution is carried out. What happens if researchers inadvertently fall prey to confirmation bias at a societal level? Would the same false results that affected the hypothetical psychopath study also affect other assumptions about human nature?
Addressing this question psychologists Joseph Henrich, Steven J. Heine and Ara Norenzayan at the University of British Columbia (where I am also located) published a paper last year in the journal Behavioral Brain Sciences. Their research documents how most of the studies that psychologists claim show human universals are really just extrapolations from a single social group, the cultural equivalent of the psychopaths in my example. As The New York Times wrote in their review:
According to the study, 68 percent of research subjects in a sample of hundreds of studies in leading psychology journals came from the United States, and 96 percent from Western industrialized nations. Of the American subjects, 67 percent were undergraduates studying psychology -- making a randomly selected American undergraduate 4,000 times likelier to be a subject than a random non-Westerner.
The subpopulation that Henrich and colleagues found to be overrepresented are entirely WEIRD (Western, Educated, Industrialized, Rich, and Democratic) societies. While it's bad enough that WEIRD American undergraduates are serving as our model for human behavior, what their paper goes on to document should be of concern to all behavioral and cognitive researchers (particularly those whose work focuses on human evolutionary explanations).
When these affluent American and non-Western populations are compared there are important differences in domains as seemingly unrelated as visual perception, fairness, cooperation, spatial reasoning, moral reasoning, reasoning styles, and even the heritability of IQ. In all cases American undergraduates didn't simply differ, they differed substantially. Nevertheless, they form the basis of most researchers' assumptions about human nature even though, as Henrich and colleagues conclude, "this particular subpopulation is highly unrepresentative of the species."
To highlight one domain in which American undergraduates differ from most other populations in the world consider a neutral category like visual perception. Looking at the figure below, which horizontal line, "a" or "b", would you estimate is longer?
If you chose "b" than you are in line with a substantial number of Americans (both undergraduates and children) who chose the same one. In fact, both lines are identical in length. This has become known as the Mller-Lyer Illusion, named after the German psychiatrist Franz Carl Mller-Lyer who first discovered it in 1889. However, if you show the same two lines to people in many non-Western societies (particularly hunter-gatherer societies) they will be more likely to identify the two lines as identical. In a series of cross-cultural experiments in 1966 psychologist Marshall H. Segall manipulated the length of line "a" until it reached the point where respondents reported that the two were identical in length. The results of these experiments can be seen in the graph below.
The vertical column represents the Point of Subjective Equality (PSE), or how long line "a" had to be before respondents said they were the same length. In other words, PSE is a measure of how effective the illusion is for different populations. As the graph indicates, Americans (labeled as "Evanston" for where Segall tested undergraduates at Northwestern University in Illinois) were the population most fooled by this illusion and required line "a" to be an average of one-fifth longer than line "b" for both to be perceived as equal. They were followed by white South Africans from Johannesburg. In contrast, the San foragers of the Kalahari were not affected by the illusion while most other societies in the study were only marginally affected.
Why would Americans be so susceptible to this illusion? Our environment. Most Americans are raised in a society where horizontal lines and sharp corners make up much of modern architecture. The brains of American children (and, presumably, most children in highly industrialized countries) have adapted to make optical calibrations as a result of their unique environment. The San and many other small-scale forager or horticultural societies don't grow up in a manufactured environment so their brains are unaffected by such illusions.
A similar difference can be found in what psychologists call "folkbiological reasoning." Cognitive scientists testing children drawn from U.S. urban centers (where most universities are located) have developed an influential developmental theory suggesting that there is a cognitive shift that takes place between ages 7 and 10. As Henrich and colleagues state in their paper:
Before age 7, urban children reason about biological phenomena by analogy to, and by extension from, humans. Between ages 7 and 10, urban children undergo a conceptual shift to the adult pattern of viewing humans as one animal among many.
This shift has been considered a process that all human children go through. The problem with this reasoning, Henrich points out, is that it only applies to one subset of children: those who live in urban environments. Similar cognitive tests of children in Native American communities in Wisconsin and among the Yukatek Maya communities in Mexico showed none of the empirical patterns that the American urban children displayed. The answer, of course, is that urban children grow up in an impoverished environment where they will rarely, if ever, interact with animals other than humans (with the occasional dog or cat kept as a pet). This is a very different environment from many non-Western societies, and certainly from the one our remote ancestors lived in.
As a result, the "unnatural" environment of these WEIRD children resulted in anthropocentric assumptions about the natural world until they were taught differently by teachers or from television (though I often wonder how an increased exposure to nature when they're young might influence adult attitudes about the importance of environmental issues). Given this, as Henrich points out, it makes as much sense to use urban children in studies of human cognition as it would to study "normal" physical growth in malnourished children. Because the psychologists who carried out these studies likely grew up in an urban environment themselves (rural students are significantly less likely to attend graduate school, particularly at top-ranking institutions) the confirmation bias of such studies are perpetuated. It's almost as if psychopaths were conducting research on themselves and claiming their results were universal.
Of course, there is one important difference between psychopaths and American society. Psychopathy, and Anti-Social Personality Disorder more generally, is a diagnosed mental disorder that has a partial basis in genetics, not just the environment. Nevertheless, the confirmation bias that exists in many psychological studies represents a distortion of reality that has just as much potential to be passed on to subsequent generations.
The fact that empirical differences exist on identical psychological studies when replicated cross-culturally should make evolutionary researchers take caution (especially Evolutionary Psychologists who are most guilty of essentializing these studies). What Henrich and colleagues have called for is a renewed effort to conduct similar cross-cultural research before making grand claims about the species as a whole. At the very least it means that researchers and science journalists alike should be careful not to perpetuate ideas that appeal to their own beliefs but which may have no basis in other societies. To do otherwise would be to confuse our own reflection in a hall of mirrors with a crowd of people making identical movements. That would clearly be psychotic.
This post originally appeared at the Public Library of Science (PLoS) Blogs.
Henrich, J., Heine, S., & Norenzayan, A. (2010). The weirdest people in the world? Behavioral and Brain Sciences, 33 (2-3), 61-83 DOI: 10.1017/S0140525X0999152X