Spatial Discrimination in the Virtual Morris Water Task: The Influence of View Similarity

Previous research in the Morris Water Task (MWT) and Virtual Morris Water Task (VMWT) examined how distal cues, or cues placed outside the pool, enabled an organism to orient themselves to navigate towards the hidden platform (Kolb & Whishaw, 2009). Recently Cheng and colleagues (2008) have created a model that predicts navigational performance based upon the similarity of views of the distal cues encountered within an environment. This model has successfully predicted performance for rats and simulated agents within a rectangular environment. The present study evaluates the view based similarity model with human subjects in the context of the VMWT. Three different distal cue configurations were tested that varied in value of pixel difference. In Experiment 1 subjects were required to navigate to a hidden platform in the VMWT and their behavioral performance corresponded to the pixel difference predictions of the model. In Experiment 2 subjects were reinforced for a correct choice between two visible platforms. Finally, in Experiment 3, the pixel comparison model was implemented to determine if it could predict the preference for directional responding to the relative location in previous virtual poolshift experiments (Hamilton et al., 2009). The predictions of the model correspond with behavioral performance in VMWT hidden platform paradigms (Experiment 1), but it made inaccurate predictions for explicit discrimination tasks (Experiment 2) and VMWT hidden platform tasks once the pool was moved within the room (Experiment 3).Masters of ScienceMastersUniversity of New Mexico. Dept. of PsychologyHamilton, DerekCaudell, TomVowles, Kevi