An exploration of the verbal encodability hypothesis for sex differences in the digit-symbol (symbol-digit) test

Two experiments were conducted to allow a test of Estes' verbal encodability hypothesis regarding female Digit-Symbol test superiority. Experiment 1 demonstrated significant differences in the relative encodability of various sets of symbols both in the time required to produce verbal associates and in quantitative measures of the uncertainty regarding what associate would be produced. Experiment 2 utilized these symbols to assess the stability of female superiority on a Symbol-Digit task across levels of verbal encodability. Predictions were that females should perform significantly better than males on tests with symbols that are easily encoded verbally, but there should be no significant sex difference when symbols are difficult to encode verbally. Significant effects for sex, task, and their interaction were found. However, the verbal encodability hypothesis is rejected for one favoring perceptual speed, due to the pattern of female superiority on the tests, and the nature of intrasex correlations. Results also show that traditional female supremacy on the Digit-Symbol carries over to the Symbol-Digit format when easily discriminable symbols are used.