Spatial cognitive implications of user interfaces in virtual reality and route guidance

Abstract

<p>The relationship between spatial learning and technology is becoming more intimately intertwined. This dissertation explores that relationship with multiple technologies and multiple types of spatial knowledge. With virtual reality, teleporting is commonly used to explore large-scale virtual environments when users are limited by the tracked physical space. Past work has shown that locomotion interfaces such as teleporting have spatial cognitive costs associated with the lack of accompanying self-motion cues for small-to-medium scale movement in virtual environments, but less is known about whether the spatial cognitive costs extend to learning a large-scale virtual environment. Experiment 1 (Chapter 2) evaluates whether rotational self-motion cues teleporting interfaces impact spatial learning for large-scale virtual environments. using two measures of survey learning (an object-to-object pointing task and map drawing task). Results indicate that access to rotational self-motion cues when teleporting led to more accurate survey representations of large-scale virtual environments. Therefore, virtual reality developers should strongly consider the benefits of rotational self-motion cues when creating locomotion interfaces. For Experiments 2 and 3 (Chapter 3), previous work has demonstrated that repeatedly using GPS route guidance reliably diminishes route learning. Memory research has shown that recalling information (i.e., testing) significantly improves retention of that information when compared to restudying the same information. Similarly, memory retrieval of routes during learning may be advantageous for long-term retention compared to following route guidance using a GPS. However, whether such a benefit would occur for route learning is not clear because the benefits of testing have primarily been explored with verbal materials. Experiments 2 and 3 explore whether retrieving routes from memory during learning enhance route knowledge of a large-scale virtual city using a driving simulator compared to learning a route by repeatedly following GPS route guidance. Results from both experiments demonstrated that there was no difference in performance between testing and repeatedly following route guidance at final test, but further analysis revealed that in the testing condition, a large proportion of errors produced during learning was also repeated at final test. The experiments described here not only expand the current knowledge regarding the intersection of technology and spatial learning, but also underscore the importance of evaluating applications of spatial cognitive theory across a range of applied domains.</p>