The influence of the design of mental rotation trials on performance and possible differences between sexes: A theoretical review and experimental investigation.

Abstract

Sex differences in mental rotation performance are one of the largest in cognitive psychology. Men outperform women by up to 1 SD in psychometric mental rotation tests, but it is often neglected that there are no or only small sex differences for chronometric tests. As both tests are supposed to measure the same ability, we suspect some features of the tests themselves to affect sex differences in performance. Following a theoretical review of the test features, we evaluate the effects of the number of possible answer alternatives, whether they are presented as pairwise mirrored, and their interaction on sex differences in mental rotation performance. In an online experiment, 838 German-speaking participants, 421 women, 417 men, M_age =  42.58 (SD = 12.54) years, solved four blocks of mental rotation trials with two or eight alternatives, which were either pairwise mirrored or not. The results show that that the overall performance was lower for more alternatives and for mixed alternatives but not for their interaction. We could not determine explanations for sex differences as we did not observe meaningful sex differences at all. Possible reasons include the differences between men and women in age and education. This study suggests that the differences between tests affect performance. Sex differences, however, need more investigation, including possible effects and interactions of the test design, education, and age.