Who developmental psychology was built on

Joseph Henrich, Steven Heine and Ara Norenzayan published an article surveying the origin of participants in studies published in the major psychology journals. They coined the acronym WEIRD for Western, educated, industrialised, rich and democratic populations.

Their finding is twofold: those populations supply the overwhelming majority of participants, and they are often atypical on the dimensions measured, including dimensions assumed universal such as visual perception. The article draws in particular on classic optical illusions, whose strength varies across the populations tested, to show that even processes deemed elementary do not escape that variation.

The field is particularly concerned because published developmental norms serve as clinical references. A norm established on a narrow population is used to situate children from other contexts.

Comparative work on motor milestones showed timing differences by carrying and floor-time practices, which led the World Health Organization to build its references on a multi-country sample.

The imbalance stems not from any particular negligence on the part of researchers but from the geography of resources: laboratories, funding and journals are for the most part located in the countries concerned, and the easiest recruitment is done close by.

Broadening samples is not merely a matter of recruiting elsewhere. Instruments must be translated, adapted and revalidated, failing which the comparison mainly measures the quality of the translation. A question about time spent playing with a child, for instance, does not refer to the same reality where children spend most of their days among peers.

Collaborative laboratory networks, which appeared in the 2010s, allow the same protocol to be run at dozens of sites. They are the main corrective tool, and they remain concentrated in countries that can fund them.

Excessive caution would mean giving up on generalisation altogether. Several findings in the field, such as the sequence of motor milestones or sleep consolidation, are found in every context studied.

The question is therefore not whether universals exist, but which ones, and the work consists in distinguishing them from local regularities taken for laws. A local regularity is not thereby without interest; it simply ceases to be a property of the species and becomes that of a context, which changes how it can be used.

The section describing the sample is the first to read, before the results. It states the country, the recruitment method and the parents' education level, information that bounds the reach of what follows.

Recruitment by notice at a university recruits particular families, which does not invalidate the study but restricts its generalisation, including within the same country.