Which cognitive subtests show the largest and most consistent sex differences across cultures?

1. Mental rotation: the clearest cross-cultural winner

Mental-rotation tasks—rotating 3D objects in the mind’s eye—produce the largest and most reliably cross-cultural male advantage reported in the literature, with moderate-to-large effect sizes that persist after controlling for education and background in many studies ^{[1] [7] [3]}. Meta-analyses and reviews repeatedly single out mental rotation as one of the largest sex differences in psychology and document its presence across dozens of countries and diverse populations ^{[3] [7]}. At the same time, authors emphasize that practice, academic background, and short-term training substantially reduce the gap, meaning the difference is robust but plastic ^[1].

2. Verbal memory and fluency: smaller but consistent female edges

Multiple cross-cultural studies report female advantages on Story Recall, word fluency, digit-symbol tasks, and reading measures—effects that emerge in adolescence and adulthood and are replicated in samples from the U.S., Japan and other nations ^{[4] [5] [2]}. These female advantages are generally smaller than the mental-rotation male advantage (often small-to-moderate in size) and can be magnified or attenuated by schooling, literacy practices, and socio-cultural context ^{[2] [6]}. Large international assessments such as PISA find girls outperform boys in reading by roughly 0.2–0.6 standard deviations across many countries, indicating consistency but also cross-national variability ^{[2] [6]}.

3. Where results diverge: math, science literacy and the tails

Sex differences in mathematics and science are far less consistent across cultures: national testing shows small average gaps and considerable cross-country heterogeneity that correlate with measures of gender equity, wealth and cultural factors ^{[8] [6]}. Importantly, some research finds sex differences are more pronounced at distributional extremes—differences in high- or low-performing tails—so average effects can mask important variance relevant to STEM representation ^{[8] [9]}.

4. The g-factor and subtest specificity: not a global intelligence gap

Large studies using many subtests show negligible sex differences in general intelligence (g) while revealing specific subtest patterns—male advantages on some spatial subtests and female advantages on several verbal and memory subtests—indicating that sex differences are localized rather than reflective of global cognitive superiority ^{[9] [10]}. This distinction guards against over-generalizing subtest effects into claims about overall intelligence, a point emphasized in recent reviews ^{[10] [11]}.

5. Causes, caveats, and implicit agendas in interpretation

Researchers propose multifactorial causes—biological (hormonal, neural), developmental, educational and cultural—that interact to produce observed patterns, and many authors stress that sociocultural modulation is large enough to challenge simple biological determinism ^{[1] [3] [11]}. Policy and ideological agendas can shape which findings are amplified: emphasizing stable biological gaps supports deterministic narratives about sex roles, while highlighting plasticity and cross-cultural variance supports educational and equity interventions ^{[6] [11]}. Methodological issues—test choice, scoring, age, and sampling—also influence which subtests show the strongest effects ^{[7] [9]}.

6. Bottom line for readers seeking reliable signals

If the question is which cognitive subtests show the largest and most consistent sex differences across cultures, the short answer is: mental rotation and closely related 3D spatial-visualization tasks show the largest, most reproducible male advantage, while several verbal-memory and fluency tasks (story recall, digit-symbol/processing-speed, word fluency, and reading) show smaller but broadly consistent female advantages; differences in other domains (math, science, working memory) are smaller, context-dependent, and often vary markedly by country, education, and methodology ^{[1] [4] [2] [6] [8]}. Where the literature does not settle the mechanism fully, it is explicit that both biology and culture play interacting roles, and that training and environmental change can substantially alter performance gaps ^{[1] [11]}.