Purpose of the study To test whether the so-called eduPGS (polygenic predictive indicators of education/intelligence) predict the level of general intelligence (g) independently of: genetic background (ancestry), parental education, skin color (color phenotype). Main results Relationship between European ancestry and intelligence (g): Hispanic Americans: r = 0.30, N = 506 African Americans: r = 0.084, N = 2179 European-African Americans: r = 0.26, N = 228 EduPGS predicts g regardless of ancestry: Hispanics: B = 0.175 Europeans: B = 0.230 European-Africans: B = 0.215 Africans: B = 0.126 The association between eduPGS and intelligence (g) persists after accounting for: skin color (genetically estimated), parental education, self-identified race/ethnicity (SIRE). Additional results Correlation between European ancestry and g: positive. Correlation between skin color and g: negative (darker color, lower g), but effect weaker and not significant after controlling for ancestry. EduPGS correlates most strongly with g in the most "g-dependent" subtests (Jensen effect). Among Hispanic Americans, those identifying as European had an average of 81% European ancestry and 29% African. Causal vs confounding models The study does not resolve whether the correlation between ancestry and g is due to: real genetic differences (causal scenario), or distortions related to sample selection and population structure (confounding scenario). In both cases, eduPGS effects are statistically significant, but the causal interpretation remains uncertain. Differences in mean IQ (g) - standardized scale scores Europeans: IQ = 100.0 Hispanics: IQ = 91.63 → d = 0.56 African Americans: IQ = 85.27 → d = 0.98 European-Africans: IQ = 97.9 → d = 0.14 The authors' conclusions EduPGS predicts intelligence in American populations regardless of skin color and social status. Differences between groups are partially explained by the EduPGS, but the source of these differences remains uncertain. Further research is needed to resolve whether and to what extent the differences are causal.