Measures of occupational status in a special population

Previous research has shown that indirect measures of occupational status such as the Duncan Socioeconomic Index (SEI) are more valid than direct measures of occupational prestige, such as those developed by the National Opinion Research Center and by Treiman, for analyses of social mobility. The Professional, Technical, and Kindred (PTK) occupations available for the SEI are a biased sample of all PTK occupations. Do findings on the validity of the SEI pertaining to the general population hold for special populations, such as college graduates? Confirmatory factor analysis comparing the validities of four measures of occupational SES—the SEI, a 1960 counterpart of the SEI developed by Siegel and based on all Census occupations, the NORC prestige scale, and the Treiman international prestige scale—showed that the indirect measure developed by Siegel was clearly more valid than the other three measures. Since these findings indicate that the predictors in a regression equation are more valid than the criterion, there is some question as to what the criterion should really be.