Sex Differences in Verbal Reasoning Are Mediated by Sex Differences in Spatial Ability

Article excerpt

There are hundreds of studies about sex differences in cognitive abilities (see Halpern, 2000, for a comprehensive review). The huge number of available databases facilitated the publication of several key meta-analyses from which some robust conclusions have been extracted. Hyde (1981) published the first meta-analysis from the data summarized by Maccoby and Jacklin (1974) in their classic book, The psychology of sex differences. The results showed that boys outperform girls in overall spatial and quantitative abilities, while the later outperform the former in overall verbal ability. Some years later, Hyde and Linn (1988) found in their meta-analysis that females outperform males in overall verbal ability. The meta-analysis reported by Hyde, Fennema, and Lamon (1990) found a male advantage in overall quantitative ability, but those researchers noted that quantitative items are frequently expressed spatially. The meta-analysis by Linn and Petersen (1985) reported a male advantage in spatial rotation, spatial relations, and visualization. Voyer, Voyer, and Bryden (1995) have replicated those findings. Finally, Feingold (1988) and Lynn (1999) observed a male advantage in reasoning ability. Therefore, previous research shows that the greater male advantage can be found in overall spatial ability (Halpern, 2000; Hedges & Nowell, 1995).

Some researchers have investigated sex differences in so-called dynamic spatial tests (Colom, Contreras, Botella, & Santacreu, 2002; Contreras, Colom, Shih, Alava, & Santacreu, 2001; Law, Pellegrino, & Hunt, 1993; Sacuzzo, Craig, Johnson, & Larson, 1996). Hunt, Pellegrino, Frick, Farr, and Alderton (1988) reported a seminal study about static and dynamic spatial abilities. "Static" refers to spatial factors measured by printed tests of visualization or spatial relations (Lohman, 2000). "Dynamic" spatial tasks measure the ability to perceive and extrapolate real motion, to predict trajectories of moving objects, and to estimate arrival times of two or more objects. Those researchers supported the view that dynamic spatial performance should be preferably measured by computerized tests (Law, Pellegrino, & Hunt, 1993; Law, Pellegrino, Mitchell, Fischer, McDonald, & Hunt, 1993; Pellegrino & Hunt, 1989; Pellegrino, Hunt, Abate, & Farr, 1987).

Contreras et al. (2001) have assessed 600 university graduates (half of them females) through two dynamic spatial tests. The computerized tests require the simultaneous orientation of two moving points to a given destination. Males outperformed females even when both come from the same or different educational branches (humanities, social sciences, natural sciences, and engineering). Thus, for instance, the scores of females graduated in engineering were lower than the scores of males graduated in humanities. Therefore, sex differences in dynamic spatial tests were not conditioned by previous educational differences.

In contrast, there are contradictory findings about sex differences in verbal ability. Females outperform males in verbal fluency (Hines, 1990), synonym generation (Halpern & Wright, 1996), or reading comprehension (Hedges & Nowell, 1995). However, males outperform females in verbal analogies (Lim, 1994). Sex differences in verbal ability measures are not always favorable to females. Nevertheless, Colom, Contreras, Arend, Botella, and Santacreu (2002) have demonstrated that performance in a test of verbal reasoning based on linear syllogisms or three-term series (John is better than Peter : Peter is better than Paul :: Who is worse?) is accurately predicted from a model of human information processing based on the mental transformation of the verbal information into a mental spatial diagram or a mental model (DeSoto, London, & Handel, 1965; Johnson-Laird, 1999; Sternberg, 1980; see also Arend, Colom, Botella, Contreras, Rubio, & Santacreu, 2003). Thus, the demonstrated male advantage in overall spatial processing is a good candidate to (partly) account for their better performance in verbal reasoning tests. …