Intelligence (Article in Press)
General mental ability in South Asians: Data from three Roma (Gypsy) communities in Serbia
J. Philippe Rushton et al.
To examine whether the Roma (Gypsy) population of Serbia, like other South Asian population groups, average lower than Europeans on g, the general factor of intelligence, we tested 323 16- to 66-year-olds (111 males; 212 females) in three different communities over a two-year-period on the Raven's Colored and/or Standard Progressive Matrices and four measures of executive function. Out of the total of 60 Matrices, the Roma solved an average of 29, placing them at the 3rd percentile on 1993 U.S. norms, yielding an IQ equivalent of 70. On the executive function tests, the Roma averaged at about the level of Serbian 10-year-olds. The Matrices showed a small mean sex difference favoring males. External validity was demonstrated by correlating the scores on Matrices with measures such as cranial capacity (r = 0.13, P < 0.01), spousal similarity (r = 0.17, P < 0.05), age at birth of first child (r = 0.26, P < 0.01), number of offspring (r = − 0.20, P < 0.01), and responsible social attitudes (r = 0.10, P < 0.05). Comparisons with extant data showed that items found difficult or easy by the Roma were those found difficult or easy by White, Indian, Colored, and Black South African 14- to 16-year-olds and by Black South African undergraduates (rs = 0.90). There was no evidence of any idiosyncratic cultural effect. Instead, Roma/non-Roma differences were found to be most pronounced on g. This was shown by item-total correlations (estimates of the item's g loading), which predicted the magnitude of Roma/non-Roma differences on those same items, regardless of from which sample the item-total correlations were calculated, and by confirmatory factor analysis. The results indicate the remarkable cross-cultural generalizability of item properties across South Asians, Europeans, and sub-Saharan Africans and that these reflect g more than culturally specific ways of thinking.