Same task, different place: Developing novel simulation environments with equivalent task difficulties

Abstract

We introduce a novel framework for creating and evaluating multiple virtual reality environments (VEs) that are naturalistic and similar in navigational complexity. We developed this framework in support of a spatial-learning study using a within-subjects design. We generated three interior environments and used graph-theoretic methods to ensure similar complexity. We then developed a scavenger-hunt task that ensured participants would visit all parts of the environments. Here, we describe VE development and a user study evaluating the relative task difficulty in the environments. Our results showed that our techniques were generally successful: the average time to complete the task was similar across environments. Some participants took longer to complete the task in one of the environments, indicating room for refinement of our framework. The methods described here should be of use for future studies using VEs, especially in within-subjects design.