Convergent Trends in Black - White Test - Score Differentials in the U . S . : A Correction of Richard Lynn

Using aggregate data from the General Social Survey (GSS), 1974-1996, Lynn (1998) claims that the Black-White intelligence difference in the United States has not been narrowing over time. We replicate Lynn's analysis and challenge his conclusion by identifying several methodological problems. By analyzing changes in Black-White differences in the GSS vocabulary test across survey years, rather than birth cohorts, Lynn overlooks both the duration and the significance of the Black-White convergence. We extend our earlier intercohort analysis of GSS data through 1998 and confirm our previous finding of a very significant, long-term Black-White convergence, which is attributable to improvements in socioeconomic background and schooling among African-Americans. Even in an analysis of aggregate changes in the Black-White test score gap across survey years, when data are weighted properly to represent the U.S. population on the individual level, we find that the Black-White test score gap narrowed significantly over the period from 1974 to 1998.