Behavioral Science is just plain WEIRD.

It is not a new observation that many of psychology’s broad claims are based on biased and ethnocentric samples (Gergen 1973; Medin & Atran 2004; Norenzayan & Heine 2005; Rozin 2001; 2009; Sears 1986; Sue et al. 1999). It’s a running joke that studies more often than not demonstrate a psychology undergraduate’s motivation for beer money than the effect of the hypothesized phenomenon on behavior. The groundbreaking study by Henrich, Heine, and Norenzayan (2010), though, managed to shine some very unflattering light on this observation, turning this once hearty inside joke into a contentious punchline, one which is usually followed by some nervous laughter.

Behavioral scientists routinely publish claims about generalizability of human psychological and behavioural phenomena in the world’s top journals based on samples entirely of WEIRD participants. WEIRD, here, is a clever, attention-grabbing, almost Gladwellian coinage for people from Western, Educated, Industrialized, Rural, and Democratic societies. In fact, the paper highlights, samples usually constitute WEIRD, American undergraduate participants. These individuals, the authors argue, are not only from a single subsample unrepresentative of the entire population, but are in fact frequent outliers, rendering them “one of the worst subpopulations one could study for generalizing about Homo Sapiens” (Henrich, et al., 2010).  The study highlights the tenuous link between making cross-cultural assertions from very specific sources of data, as well as the rampancy with which this takes place in psychology and the behavioral sciences. A systematic survey of papers published between 2003 and 2007 in leading scientific journals belonging to six different subfields of psychology found that while these papers made broad claims about the human mind, they were almost exclusively based on WEIRD samples. A striking example of this, the author notes, was found in papers published in the Journal of Personality and Social Psychology – arguably one of the most important journals in this subfield – wherein 96% of the participants were WEIRD, and 68% were Americans. This means that 96% of psychological samples come from countries with only 12% of the world’s population. Moreover, two-thirds of the individuals sampled for the papers published were psychology students in Western universities. Henrich et al. jokingly suggested the journal should change its name to the Journal of Personality and Social Psychology of American Psychology Students.   

The paper reviewed findings and identified population variation in aspects of visual perception, memory, attention, fairness motivations, categorization, induction, spatial cognition, self-enhancement, moral reasoning, defensive responses to thoughts about death, and heritability estimates of IQ.  The findings stated in the original research articles were thought to be generalizable to “people” at large, but Henrich et al. maintain that the paucity in cross-cultural findings is still a hindrance to truly understanding any of these phenomena. These aspects span across social and nonsocial facets of the environment and can be argued as being ‘fundamental’ to the human experience. If the seminal experiments in each of these fields were not conducted in a WEIRD society with a WEIRD sample but instead, say by Bodo speakers of Assam, would the theoretical foci and primary results be the same? Or would they be completely different? It is scary to think that due to the dearth of comparative data across diverse populations, the theories that undergraduates at Indian Universities learn in class may not be applicable to their environment, schemas, agency and/or development.

While the paper quite brilliantly lays out the pitfalls leading to the bias towards WEIRD participants and the dangers of the same, one problem it does not quite fully examine is that there aren’t just too many WEIRD participants, but that there too many WEIRD researchers as well. How the behavioral sciences community is constituted itself is the first source of bias. The dominance of research emerging from WEIRD societies - due to better academic funding, systemic opportunities and a whole host of other reasons – robs the global community of worthwhile perspectives, critiques, and hypotheses. In the open peer-commentary section of the paper, a section truly worth scrolling down to the end of the paper and reading, Meadon and Spurrett (2010) quite aptly state that the benefits of more Western researchers are much smaller than the benefits of more non-Western researchers, “just because they are non-Western”. This would contribute not only to additional subjects but also to additional novel perspectives and ideas, less affected by WEIRD biases.

It is clear that there is a pressing need to foster academic research capacity in non-Western countries, such as India. Long-term inter-institutional collaboration, extensive student exchange programs, open-source sharing of research materials, active collaboration between academics, governments and other grant-making agencies, and a change in journal editorial and review methods to strengthen claims of generalizability should be some steps taken to build a wide, rich, and truly global understanding of humans.

Aneree Parekh