Different tasks are often used to assess spatial memory in humans compared to nonhumans. In order to bridge this paradigmatic gap, we used a within-subject design to test 61 undergraduates on three spatial memory tasks. One of these tasks, the Vanderberg 3D mental rotation task, is classically used to assess spatial memory in humans. The other two tests are virtual analogues of two tasks used classically to assess spatial memory in rodents: the Morris water task and an eight-arm radial maze. We find that males perform significantly better than females on the mental rotation task and in finding a hidden platform in the virtual Morris water task. Moreover, during a probe trial, males spend significantly more distance of their swim in the training quadrant, but males and females do not differ in navigating to a visible platform. However, for the virtual eight-arm radial maze, there is no sex difference in working memory errors, reference memory errors, or distance to find the rewards. Surprisingly, an examination of the correlations among the three tasks indicates that only mental rotation ability and Morris water task probe trial performance correlate significantly among the three tasks (i.e. there are no significant correlations with traditional measures the tasks, e.g. time or distance to completion). Hence, the Morris water task and the eight-arm radial maze do not assess spatial memory in the same manner, and even after equating factors such as motivation, stress, and motor demands, there still are procedural demands of the tasks that reinforce differential strategy selection during spatial memory. This suggests that caution should be taken when utilizing these two tasks interchangeable as tests of spatial memory.
Copyright 2003 Elsevier B.V.