I have found
this today in ScienceDirect (probably not available at your institution):
Past research has demonstrated that when spatial knowledge is acquired through real-world navigation, males typically outperform females on subsequent tests of spatial ability. However, less-consistent gender differences have been obtained in studies examining navigation through computer-simulated spaces. In the present study, male and female participants explored two large-scale virtual shopping centres. Navigation was supported by either a hand-held paper map or by a digital map that was displayed on a computer screen. Spatial knowledge was then tested in a battery of tasks including wayfinding, directional and distance estimates, and a map placement task. In the majority of comparisons made, there were significant differences favouring males. The findings indicate that like real-world learning, virtual exploration leads to significant gender differences in spatial performance.
If the experiment had tested also ability at map-folding, it would have been absolutely perfect! There's a
bit more of information, and screenshots, at one of the researchers' home page.
