The utility of mental models and visualization in training

Traditionally, learning about spatial layout has been based on text descriptions or 2-dimensional layouts, such as maps. Recent developments in technology have provided a new tool for such learning and navigation, virtual reality (VR). This study examined how people use VR in developing spatial mental models and compared that approach with other visual tools. Five types of representation were used to train people on the layout of a simple VR building. The Text condition was a narrative description of the environment. The 2D condition was bird's eye view with sequential selection of rooms. The Movie condition was a prerecorded movie displayed on a computer monitor. Passive-VR was a guided walkthrough of the prerecorded movie presented on a Head Mounted Display (HMD) with no user control. Active-VR was the prerecorded movie presentation on the HMD with user control of rotational motion. Learning was measured as number of trials to correctly identify all objects. Resultant spatial mental models were assessed by a reaction-time task indicating whether two objects were in the same or different rooms and by drawings of the objects and layout of the virtual space. Movie condition participants learned the location of items more rapidly than did participants in any of the other conditions. Movie and Active-VR reaction times were faster than the other three conditions. Speed of response was related to distance of objects along the path taken through the building, suggesting that route distance was represented in the participants' mental models. The Movie and Active-VR conditions also had the highest scores on drawing. The results suggest that learning rate may be a function of familiarity with a particular tool, whereas the character and accessibility of spatial mental models may be due to a combination of prior experience and immersion.