Human overshadowing in a virtual pool: Simple guidance is a good competitor against locale learning

In four experiments a new virtual preparation for humans of the Morris water task (VMWT) was used. Psychology students were trained to locate a platform (either visible or invisible) in the presence of four landmarks (A, B, C, D), spaced at equal intervals around the edge of the pool. At the end of training one test trial was given in the presence of one or several landmarks, without the platform, and the time the students spent in the platform quadrant was registered. Experiment 1 used an invisible platform. It was designed to see how much the students had learned either of the whole set of four landmarks or of some subset of it when searching for the platform on test. When tested with four or two landmarks (either relatively near or far from the platform), the students' performance was equivalent and significantly better than that obtained with one landmark only (either relatively near or far from the platform). In Experiment 2, for Group Experimental, the platform was visible, while for Group Control, it was invisible. On the test trial, a clear overshadowing effect was found: the Overshadowing group spent significantly less time in the platform quadrant than the Control group. A third group, Group Experimental-Slow, was subsequently added to eliminate an alternative explanation of spatial overshadowing in terms of differential experience with the landmarks during training. Finally, Experiments 3 and 4 were conducted to control for generalization decrement. The data are discussed within the growing body of evidence that suggests that the general laws of learning apply to many species, both in the spatial and temporal domains. (C) 2003 Elsevier Science (USA). All rights reserved.