WEIRD psychological science - geographical patterns of knowledge production

Psychology has argued about its WEIRD problem since 2010, when Joseph Henrich and colleagues pointed out that a discipline claiming to describe human behavior draws its participants overwhelmingly from Western, Educated, Industrialized, Rich, and Democratic countries. Lejla Džanko, Aleksandra Cisłak, and Magdalena Formanowicz at SWPS University in Warsaw asked the same question about the researchers.

The dataset

The team pulled every psychology journal indexed in Scopus, then used the Elsevier Research Products API to download the articles those journals published between 2020 and 2023. After dropping journals with no citation metrics and articles with no authorship data, 1,398 journals and 317,006 articles remained, carrying 1,273,571 authors, an average of 4.02 per paper.

Gender came from Genderize.io. Authors' first names went in; a binary classification and a probability of the guess being correct came back. The tool returned no classification for 7,361 authors, who were treated as missing data. Geography came from self-reported affiliation: the country of an author's institution, coded WEIRD if it sat in the UN's "Western European and Other States" regional group. Türkiye was excluded from that list as a poor fit with the definition used in the literature; Israel was kept.

Four Scopus metrics stood in for journal prestige: CiteScore, %Cited, SNIP, and SJR. Each was tested in a separate model with papers nested inside journals, since articles in the same journal share editorial policy and impact scores.

Who publishes

Across the four years, 72% of authors were affiliated with institutions in WEIRD countries. The prestige positions were no better: 72% of articles had a WEIRD first author, 73% a WEIRD last author.

That share moved with impact. On every metric tested, a higher-impact journal meant a higher proportion of WEIRD authors and a higher probability that the first or last author was WEIRD. The effect was strongest for SNIP and SJR, the two metrics that adjust for field citation habits and weight citations by the prestige of the citing journal.

Time pulled the other way, but weakly. The WEIRD share declined slightly from 2020 to 2023. That decline did not reach the high-impact journals, where the prestige effect ran opposite and stronger. More non-WEIRD authors are publishing in psychology. They are not publishing in its most consequential venues.

The gender result

Gender barely moved. The proportion of women authors showed no stable relationship with journal impact across the four models. Women's probability of holding the last-author position rose over the four years, and two metrics, CiteScore and SJR, showed a weak negative association between journal impact and a female last author. Otherwise, authorship looked close to balanced.

The authors argue that balance is not good news. Women are the majority in academic psychology: among psychology students, in the American Psychological Association, and in the Society for Personality and Social Psychology. Measured against those base rates, a 50/50 authorship split is a shortfall, not parity. The paper calls it representation bias.

What the authors flag

The authors are blunt about where the coding breaks. Genderize, they write, "relies on binarity and stereotypicality of names," and the WEIRD classification required judgment calls that the field's most-cited definitions disagree on. Affiliation stands in for origin without measuring it: a researcher trained in Lagos and working in Boston counts as WEIRD here. They suggest tracking how affiliation changes over time rather than fixing each author at one address.