Three-Dimensional (3D) virtual environments enable users to interact with information in a spatially realistic manner that can facilitate visualization and manipulation of information. Yet we still know little about how users navigate such environments, and in particular we know little about how past spatial experience influences navigation performance. In this study, we tested 82 participants in two experiments to examine their navigation performance in different versions of the virtual Morris Water Maze, analyzing eye fixations and individual differences in mental rotation ability. We assessed navigational competence by comparing participants' search times to locate a hidden target and their dwell times on the actual target area. In general, the results showed that past 3D video game experience was associated with better navigation performance in some, but not all, instances. We found similar sex differences as have been observed in prior research, confirming that there is a gender-differentiated uptake of distal and proximal cues, with navigators of high spatial ability being less reliant on salient proximal cues than navigators of low spatial ability -but we were also able to demonstrate how, with the provision of proximal cues, in the form of landmarks on the circumference of a virtual pool, these gender-based differences in navigational performance were significantly diminished. Given that spatial abilities have been correlated with positive educational outcomes in STEM (science, technology, engineering and mathematics) subjects, and that female students have been and remain under-represented in these subject areas despite their centrality to 21st century educational and occupational demands, understanding the factors that affect abilities to navigate virtual spaces, identifying the underlying processes that different users bring to bear when navigating 3D virtual environments, and knowing what design modifications support improvements in performance are important to the advancement of educational opportunities. This study contributes to realizing that larger goal in several respects: theoretically, it provides insights into how different types of environmental information impact 3D virtual navigation performance and how spatial learning is affected; methodologically, it demonstrates uses of novel research tools and designs to better understand virtual navigation, including the use of an eye tracker and virtual versions of the Morris maze. As 3D virtual environments proliferate, it will become increasingly important to understand how different types of cues affect navigational performance. Our work shows how manipulating the proximity and saliency of cues impacts spatial learning and goal completion.
