Abstract
Background
This study examined the effects of automated decision-support tools (DSTs) on novices’ cognitive workload in a simulated shipboard damage control environment. With the growing reliance on technology and personnel shortages, novices often take high-stakes roles without experience, making it crucial to understand DSTs' impact on novice decision-making. Prior research has shown that DSTs enhance situation awareness and performance for experienced users, but there is a gap in understanding their effects on novices who lack prior exposure to high-stakes operational environments. Cognitive workload theory suggests that the effectiveness of DSTs depends on how the cognitive resources are distributed, with different modalities influencing workload and decision efficiency in distinct ways. Virtual Reality (VR) serves as the research platform, providing an immersive setting to simulate shipboard damage control tasks and systematically examine cognitive workload and performance under varying DST modalities.
Significance
This study directly supports the Navy’s initiatives to integrate automation into training and operations by evaluating how DSTs impact novice cognitive workload in high-stakes naval damage control environments. The findings provide critical insights into optimizing DST design to ensure alignment with the cognitive capabilities of inexperienced personnel, helping to mitigate cognitive overload, improve decision-making efficiency, and enhance knowledge transfer to the trainees in naval operations. The study’s insights are valuable to communities exploring the implementation of AI-driven decision aids in dynamic operational environments.
Methods
This study employed a within-subject experimental design to assess cognitive workload and task performance in approximately 30 novice participants operating in a VR-based damage control simulation. Workload measurements include NASA-TLX ratings, physiological indicators (HRV and GSR), and objective task performance metrics. Each participant completed four experimental conditions in randomized order:
- Scenario with no DSTs.
- Scenario with visual DSTs.
- Scenario with audio DSTs.
- Scenario with audio and visual DSTs.
Results
The experiment is currently underway. Data collection and analysis will be completed by May 2025.
