works
B Facon Are correlations between cognitive abilities highest in low-IQ groups during childhood? article This paper focuses on Spearman’s law of diminishing returns which states that correlations between IQ tests decrease as the intellectual efficiency increases. In the present study, data from the national standardization sample of a French intelligence scale for children aged 4 to 9 years (Echelles Différentielles d’Efficiences Intellectuelles, forme Révisée) were examined to confirm this relationship. Each of the seven subtests of this scale was successively used to divide the sample into two IQ groups (low vs. high IQ) and correlations between the remaining six subtests were computed for each group. Fit measures of matrices revealed that correlations were not statistically different for six of the seven comparisons between low- and high-IQ participants. These results seem to indicate, at least in children aged 4 to 9 years, that lower IQ samples do not manifest a less differentiated pattern of correlations than higher IQ samples. Some theoretical implications of this finding are discussed, notably the need to envisage age as a potentially meaningful variable in research on the law of diminishing returns. © 2004 Elsevier Inc. All rights reserved.

Are correlations between cognitive abilities highest in low-IQ groups during childhood?

B Facon

Intelligence, vol. 32, no. 4, 2004, pp. 391–401

Abstract

This paper focuses on Spearman’s law of diminishing returns which states that correlations between IQ tests decrease as the intellectual efficiency increases. In the present study, data from the national standardization sample of a French intelligence scale for children aged 4 to 9 years (Echelles Différentielles d’Efficiences Intellectuelles, forme Révisée) were examined to confirm this relationship. Each of the seven subtests of this scale was successively used to divide the sample into two IQ groups (low vs. high IQ) and correlations between the remaining six subtests were computed for each group. Fit measures of matrices revealed that correlations were not statistically different for six of the seven comparisons between low- and high-IQ participants. These results seem to indicate, at least in children aged 4 to 9 years, that lower IQ samples do not manifest a less differentiated pattern of correlations than higher IQ samples. Some theoretical implications of this finding are discussed, notably the need to envisage age as a potentially meaningful variable in research on the law of diminishing returns. © 2004 Elsevier Inc. All rights reserved.

PDF

First page of PDF