Sebastian Jensen

Past research suggests that national stereotypes are largely inaccurate, unlike other demographic stereotypes. However, past research relied mainly on self-reported personality as the truth set (criterion data), which is problematic due to the reference group effect. Indeed, when an objective measure of conscientiousness was constructed using the measures such as the accuracy of public clocks, this correlated well with national stereotypes (r’s .60 to .70). No prior measurements of national stereotypes of intelligence had been reported in the literature. We surveyed a nationally representative sample of 478 Americans on Prolific to assess the stereotypes about national intelligence, and related these to estimates of national intelligence. We found that stereotypes about national IQs correlated highly with national IQ data (r = .88).

At the individual level in bivariate analysis, leftism (r = -.21, p < .001), neoconservatism (r = -0.24, p < .001), being Black (d = -0.54, p < .001), and being female (d = -0.44, p < .001) predicted lower accuracy. Crystalized ability was not a robust predictor of accuracy (r = .011, p = .86), though the quality of the measurement of crystalized ability was questionable.

Correlates of the ratings of specific countries were also assessed. Leftists tended to rate the intelligence of Americans, Russians and Israelis lower than right wingers, while they assessed the intelligence of Cuba, Uganda, and Kenya to be higher in comparison to right wingers. Blacks rated the intelligence of African countries higher (d = 1.00, p < .001) than Whites, and leftism was associated with giving higher ratings to African countries (r = .26, p < .001). Some of the analysis in this paper were preregistered.