Sex differences in problem solving.

Abstract

Nine experiments were performed to verify and extend studies on sex differences in problem solving conducted in the 1950s by Sweeney, Carey, Milton, Nakamura, and Berry. A 20-item set of problems similar to that used by Sweeney et al. was administered to over 1,000 college students using a 3-min per problem time limit. Results indicated that (a) the male advantage, averaging 35% across experiments, persists at the same level as in the 1950s; (b) prior experience with the particular word problems used is unrelated to the sex difference; (c) the sex difference extends to other word problems but not to closed-context reasoning problems or measures of general intelligence; (d) the male advantage is related to similar advantages in spatial and mathematical ability; and (e) aptitude variables dominate attitude and mathematics experience variables in accounting for the sex difference.Reviews of cognitive sex differences over the last two decades are in general agreement that males are superior to females in problem solving. It is not readily perceived, however, that this assertion rests largely on a small number of studies conducted in the 1950s by Sweeney, Milton, Carey, and Nakamura, all of whom were students of Donald Taylor during his years at Stanford University and later at Yale University. As an example of the influence of these studies, eight fairly recent reviews (

Sex Differences in Spatial Cognition, Computational Fluency, and Arithmetical Reasoning

“…The latter finding is in keeping with Jensen's (1998) and Geary's (1998) conclusions, based on reviews of the literature, that there are no mean sex differences in general intelligence (but see Rushton & Ankney, 1996). The magnitude of the male advantage on the MRT was in the range found in previous studies (Geary, 1998;Linn & Petersen, 1985;Voyer, Voyer, & Bryden, 1995), as was the male advantage on the arithmetical reasoning tests (Geary, 1996;Johnson, 1984;.…”

“…The magnitude of the relation between spatial abilities and skill at solving arithmetical word problems was somewhat smaller than that found in studies of the relation between spatial abilities and performance on the SAT-M and on algebraic word problems (Casey et al, 1997;Johnson, 1984). In the current study, the path coefficients from spatial cognition to arithmetical reasoning ranged from .13 to .19.…”

“…In the current study, the path coefficients from spatial cognition to arithmetical reasoning ranged from .13 to .19. The coefficient, based on a path analysis, between spatial abilities and SAT-M performance was .23 in the Casey et al (1995) study and the partial correlation (controlling for SAT-M performance) between spatial abilities and algebraic problem solving was .29 in Johnson's (1984) study. Differences across studies are likely to reflect differences in the types of items used to assess mathematical problem solving, as it appears that the relation between spatial abilities and mathematical problem solving abilities varies from one word problem to another (Johnson, 1984).…”

“…The coefficient, based on a path analysis, between spatial abilities and SAT-M performance was .23 in the Casey et al (1995) study and the partial correlation (controlling for SAT-M performance) between spatial abilities and algebraic problem solving was .29 in Johnson's (1984) study. Differences across studies are likely to reflect differences in the types of items used to assess mathematical problem solving, as it appears that the relation between spatial abilities and mathematical problem solving abilities varies from one word problem to another (Johnson, 1984). Presumably word problems differ in the extent to which spatial representations will facilitate performance (i.e., reduce problem solving errors) and in the ease with which the quantitative relations conveyed in the problems can be translated into spatial representations (Geary, 1996;Lewis, 1989).…”

“…1, model 4 involved the estimation of a directed path from the Spatial Cognition latent variable to the Arithmetical Reasoning latent variable, based on the hypothesis that spatial abilities contribute to individual differences in the ability to solve mathematical word problems (Geary, 1996;Johnson, 1984;Lewis, 1989). The estimation of Model 4 resulted in a marginally significant improvement in model fit, ⌬ 2 (1) ϭ 3.31, p Ͻ .08.…”