Repository logo

Sex differences in general intelligence: a psychometric investigation of group differences in mean and variability as measured by the Raven's Standard Progressive Matrices

Change log


Savage-McGlynn, Emily 


Researchers and the general public alike continue to debate ‘which is the smarter sex?’ Research to date suggests that males outperform females, females outperform males, while others find no differences in mean or variance. These inconsistent results are thought to occur for two reasons. First, studies rely on opportunity samples rather than samples that represent the general population. Second, researchers have not availed themselves of advances in psychometrics that allow for identification of bias in test items and the reliable evaluation of group differences. This dissertation addresses these two identified needs in the literature.

Using a large representative U.K. sample, 926 seven to 18 year olds were assessed with the Raven’s Standard Progressive Matrices Plus (SPM+), a measure considered to be one of the best measures of general intelligence. In assessing a one-factor model of general intelligence, four research aims were addressed. First, confirmatory factor analyses and assessment of measurement invariance revealed that the SPM+ is not biased to either sex. Second, multiple group confirmatory factor analyses revealed there to be no significant differences between males and females in either mean or variance. Third, analyses revealed no significant sex differences in mean or variability in younger or older participants. Finally, method effects of Gestalt and Visuospatial answering strategies explained some of the residual variance in the model. For the overall sample, males were significantly disadvantaged by the visuospatial element of some of the items. For older participants, the influence of the methods effects was equivalent.

It can generally be concluded that there are no significant sex differences in mean or variability on the SPM+ suggesting that there is no sex difference in general intelligence. Future research should employ representative samples and robust statistical methodologies to assess sex differences on the Raven’s from a multiple factor perspective.






Doctor of Philosophy (PhD)

Awarding Institution

University of Cambridge