WEIRD samples
Most behavioral findings come from a narrow, globally unrepresentative slice of humanity.
What it means
WEIRD is an acronym — Western, Educated, Industrialized, Rich, and Democratic — that flags the fact that the overwhelming majority of participants in behavioral and psychological research are drawn from societies that are statistical outliers on a global scale, often university undergraduates from a handful of countries. The critique is that such samples are routinely treated as a stand-in for humanity in general, even though cross-cultural research shows that WEIRD populations are unusual on many dimensions thought to be basic, including visual perception, spatial reasoning, fairness and cooperation norms, moral intuitions, and styles of self-construal. The mechanism of the problem is sampling: convenient, accessible subjects are not a random draw from the human population, so findings calibrated on them may not generalize, making this a major threat to external validity. The concern does not invalidate the findings within their populations, but it cautions strongly against universal claims about 'human nature' built on a narrow base. The remedy is broader, more diverse and cross-cultural sampling. It matters because behavioral science aspires to describe people in general, yet much of its evidence rests on some of the least typical people on Earth.
Examples
Susceptibility to classic visual illusions and notions of fairness differ markedly between WEIRD and non-WEIRD populations, undermining claims that lab findings are universal.
A team tests its new app only on London-based students, then ships worldwide and finds that users elsewhere read the swipe gestures and the icons completely differently.
A textbook chapter titled 'how humans choose' rests almost entirely on studies of psychology undergraduates paid course credit in a handful of rich countries.
First described in Henrich, Heine & Norenzayan (2010).