Sex Differences in Math Problem Solving as a Function of Gender-Specific Item Content

Abstract

Among the hypotheses advanced to explain sex differences in math problem solving is the assertion that males are assisted by male-familiar item content and females by female-familiar item content. A 15-item test with parallel forms was constructed. Each form contained 5 male-familiar items, 5 female-familiar items, and 5 neutrally worded items. Male-familiar items on Form A were rephrased to construct the female-familiar items of Form B. Similarly, the female-familiar items of Form A were reconstructed as the male-familiar items of Form B. Neutral items were identically worded on the two forms. Thus, the items were equal with respect to the operations needed to solve them as well as the complexity of units and numerical relationships within them. Males outperformed females in all item conditions. The differences in observed summation scores were statistically significant between the two groups. Gender-familiar wording was unrelated to the outcome. The hypothesis that biased item wording contributes to gender differences in math problem solving was not supported.