Using a version of Kelly's Role Construct Repertory Grid Test, experiments were conducted amongst samples of adolescents to investigate their modes of construing value concepts. Analysis concentrated on the definition of levels of abstraction and insight associated with individual perceptions of such concepts, a hierarchical structure involving three levels of abstraction being indicated. Results suggested that measured intelligence and social background were significantly correlated with the use of higher levels of abstraction, these correlations remaining significant when the correlation between measured intelligence and social background had been taken into account. No significant relationship was found between the sex of subjects and the levels of abstraction.

Summary. Four indices of achievement relative to a measure of general intelligence were constructed for a sample of 178 adolescents. Measures of achievement were taken from tests in reading comprehension and mathematics from an English essay and from the results of public examinations at 16+. Relative achievement measures were derived from the regression residuals of the four separate achievement scores against scores on the AH4 test of general intelligence. Using reliability estimates for two achievement tests, reading comprehension and mathematics, and for the intelligence test, it was calculated that 69 per cent of the variance in the regression residual for reading comprehension, and 56 per cent in that for mathematics, were independent of unreliabilities in the tests. Analysis of the relationships between the four regression residuals indicated a single major component on which all the residuals were highly loaded, but the low correlation found between essay and mathematics with intelligence controlled suggested the involvement of two independent skills. Correlations with an additional set of variables indicated that pupils from poorer social backgrounds, with poorer records of attendance and poorer ratings of conduct in school obtained lower scores relative to their intelligence test scores for all the achievement measures; boys tended to obtain lower scores than girls except in mathematics where there was no significant sex difference.

No abstract