Visual attention in supervisory control: Examining the effects of mental models, information bandwidth, and...

Modern aviation environments require operators to monitor systems with extensive automation. This research examines how mental models of automated systems affect visual attention allocation during supervisory control tasks such as remote aircraft operations. We hypothesized that advanced mental models would increase attention to task-relevant information and that high-bandwidth information would draw more attention only when task-relevant. Fifty-five university students completed simulated supervisory control scenarios after receiving training with or without explicit automation functionality information (Advanced or Basic Mental Model groups), with eye movements recorded. The Advanced group allocated more attention to supplemental data displays than the Basic group. Unexpectedly, low-bandwidth information received more attention than high-bandwidth when deemed task-relevant, suggesting that contextual cues, rather than bandwidth alone, may shape expectancy in this task environment. Results suggest the value parameter in the Salience, Effort, Expectancy, Value model is sensitive to mental model differences, with a pattern most consistent with a multiplicative relationship between expectancy and value parameters. Computational modeling showed that Salience, Effort, Expectancy, Value models with multiplicative terms and expectancy coefficients based on discrete event rates provided the most accurate predictions. Practically, these findings indicate that simple training to refine mental models can alter visual scanning behavior in supervisory control tasks. (PsycInfo Database Record (c) 2026 APA, all rights reserved).