Influence of educational attainment upon tests of intelligence
AbstractBurt's experiment is repeated and his conclusions critically examined. There were some extensions. In addition to Burt's reasoning tests, the Binet-Simon tests and school marks, there were used the Dearborn group tests and marks on individual subjects of instruction. The subjects were 77 pupils, all of whom were nine years of age and scattered from Grade II to Grade V. Correlations were then computed between the several variables. For example, the correlation of the Binet test with school marks was .77, while Burt found with the same criterion .72. Burt found the correlation between the Burt test and school work, with the Binet test constant, to be––.07; Freeman finds the corresponding figure to be + .31. The Binet test correlates more highly with school marks than does the Burt test. The average Burt I.Q. of the group of 77 pupils mentioned above was 89; on the Binet it was 103. The investigation does not confirm Burt's findings that Burt's reasoning test is a test of pure intelligence, or that the Binet scores depend largely on school attainment.