Longitudinal correlation analysis of standing height and intelligence

Intercorrelations of 10 successive years of measurement of height and intelligence are presented for separate samples of girls and boys. These correlations are based on data originally gathered and published by Dearborn, Rothney, and Shuttleworth as the Harvard Growth Study. Sample size varies from correlation to correlation, but most of those for girls are based on samples of 500-700 and those for boys on samples of 400-500. The intercorrelations of each of the 2 variables over 10 occasions do not differ appreciably by sex, but there are significant differences between the sexes in the cross-correlations. For the sample of girls there is clear evidence that individual differences in height at 8 and 9 anticipate later individual differences in intelligence. Correlations of early height with intelligence at 11 and 12 are especially high (.40). There is little evidence for similar anticipation of intelligence by height for boys. Correlates for both height and intelligence are found in socioeconomic status, ethnicity, and age of first menstruation for girls. Only the last of these contributes to the explanation of the changes in the cross-correlations with age. Analyses of sitting-height correlations with intelligence indicate that length of the long bones of the legs is also related to the observed pattern of correlations.