Abstract
Considerable evidence shows that people have difficulty maintaining orientation in virtual environments. This difficulty is usually attributed to poor idiothetic cues, such as the absence of proprioception. The absence of proprioceptive cues makes a strong argument against the use of a joystick interface for locomotion. The importance of full physical movement for navigation has also recently been confirmed (Ruddle and Lessels, 2006), where subjects performed a navigational task better when they walked freely rather than when they could only physically rotate or only move virtually.
Our experiment replicates the experiment of Ruddle and Lessels but under different conditions. Here all conditions are conducted using a head-mounted display, whereas Ruddle and Lessels mixed display types. Our environment contains no environmental cues to geometry, as all landmarks are either randomly placed and oriented, or absent, whereas the Ruddle and Lessels environment included a simulated rectangular room that was always visible. People are sensitive to environmental geometry, but the effect on navigation is an active area of research (Kelly et al., 2008), thus our environment omitted them.
In this experiment, subjects (N=12) locomoted through an environment in one of three ways: they walked, they used the joystick to translate while physically rotating their bodies to change orientation, or they used a joystick to both translate and rotate with no physical movement occurring.
A within-subjects design found that subjects were marginally better in the walking condition than in other conditions (F(1,11) = 2.88, p=.07). Subjects were significantly slower in the joystick condition than in other conditions (F(1,1)=5.44, p=.01). Subjects traveled significantly less distance in completing the task in the walking condition than in other conditions (F(1,11)=4.28, p=.03). In general, we conclude that walking seems a better method for locomotion in virtual environments than locomoting with a joystick.
The authors thank Jon Kelley and John Rieser for helpful discussions.