Testing the cross-racial generality of Spearman's hypothesis in two samples

Spearman's hypothesis states that racial differences in IQ between Blacks (B) and Whites (W) are due primarily to differences in the g factor. This hypothesis is often confirmed, but it is less certain whether it generalizes to other races. We therefore tested its cross-racial generality by comparing American subjects of European descent (W) to American Hispanics (H) in two different databases. The first [Centers for Disease Control (1988). Health status of Vietnam veterans. Journal of the American Medical Association 259, 2701–2719; Centers for Disease Control (1989). Health status of Vietnam veterans: Vol IV. Psychological and neuropsychological evaluation. Atlanta, Georgia: Center for Environmental Health and Injury Control] contains 4462 middle-aged Armed Services Veterans males, and the second database (NLSY1979) holds 11,625 young male and female adults. Both samples are fairly representative of the general American population.Race differences in general intelligence g were calculated and vectors of test scores were correlated with the vectors of the tests' g loadings, following Jensen [Jensen, A. R. (1998). The g factor. Westport, CT: Praeger].W scored about 0.8 S.D. above H. The racial difference on the tests correlated significantly with the g-loadings of the tests in the VES database, but less so in the NLSY database.We therefore conclude that the present study supports, but does not unequivocally verify, the cross-racial generality of the Spearman's hypothesis.