Effects of Field of View and Visual Complexity on Virtual Reality Training Effectiveness for a Visual Scanning Task

Journal Article · Thu Feb 12 23:00:00 EST 2015 · IEEE Transactions on Visualization and Computer Graphics

DOI: https://doi.org/10.1109/TVCG.2015.2403312 · OSTI ID:1185475

Ragan, Eric D. ^[1]; Bowman, Doug A. ^[2]; Kopper, Regis ^[3]; Stinson, Cheryl ^[4]; Scerbo, Siroberto ^[2]; McMahan, Ryan P. ^[5]

---

1. Oak Ridge National Lab. (ORNL), Oak Ridge, TN (United States). Cyber and Information Security Group
2. Virginia Polytechnic Inst. and State Univ. (Virginia Tech), Blacksburg, VA (United States). Dept. of Computer Science
3. Duke Univ., Durham, NC (United States)
4. Precision Nutrition, Toronto, ON (Canada)
5. Univ. of Texas-Dallas, Richardson, TX (United States)

Virtual reality training systems are commonly used in a variety of domains, and it is important to understand how the realism of a training simulation influences training effectiveness. The paper presents a framework for evaluating the effects of virtual reality fidelity based on an analysis of a simulation’s display, interaction, and scenario components. Following this framework, we conducted a controlled experiment to test the effects of fidelity on training effectiveness for a visual scanning task. The experiment varied the levels of field of view and visual realism during a training phase and then evaluated scanning performance with the simulator’s highest level of fidelity. To assess scanning performance, we measured target detection and adherence to a prescribed strategy. The results show that both field of view and visual realism significantly affected target detection during training; higher field of view led to better performance and higher visual realism worsened performance. Additionally, the level of visual realism during training significantly affected learning of the prescribed visual scanning strategy, providing evidence that high visual realism was important for learning the technique. The results also demonstrate that task performance during training was not always a sufficient measure of mastery of an instructed technique. That is, if learning a prescribed strategy or skill is the goal of a training exercise, performance in a simulation may not be an appropriate indicator of effectiveness outside of training—evaluation in a more realistic setting may be necessary.

• Research Organization: Oak Ridge National Laboratory (ORNL), Oak Ridge, TN (United States)
• Sponsoring Organization: USDOE; Office ofNaval Research
• Grant/Contract Number: AC05-00OR22725
• OSTI ID: 1185475
• Journal Information: IEEE Transactions on Visualization and Computer Graphics, Journal Name: IEEE Transactions on Visualization and Computer Graphics Journal Issue: 7 Vol. 21; ISSN 1077-2626
• Publisher: IEEECopyright Statement
• Country of Publication: United States
• Language: English