Estimation of Subjectively Reported Trust, Mental Workload, and Situation Awareness Using Unobtrusive Measures

Area of Science:

• Human-Computer Interaction
• Cognitive Engineering
• Human Factors in Aviation and Spaceflight

Background:

• Subjective questionnaires are standard for assessing human cognitive states like trust, mental workload, and situation awareness (TWSA).
• These questionnaires are obtrusive and often impractical for use during operational tasks.
• Embedded measures, derived from operator actions, offer a potential unobtrusive alternative for TWSA estimation but require systematic investigation.

Purpose of the Study:

• To systematically investigate embedded measures for estimating trust, mental workload, and situation awareness (TWSA).
• To develop and evaluate statistical models that leverage embedded measures for accurate cognitive state estimation.
• To determine the utility of embedded measures in operational contexts where traditional questionnaires are infeasible.

Main Methods:

• Fifteen participants performed spaceflight-relevant tasks using a simulated autonomous system.
• Embedded measures of TWSA were collected during task performance.
• Statistical models were developed and refined using embedded measures, task information, and participant background data.

Main Results:

• Embedded measures often correlated with intuitive changes in reported TWSA.
• Embedded measures alone were insufficient for precise cognitive state capture.
• Models combining embedded measures with task or participant data achieved good descriptive fit and accurate TWSA prediction.

Conclusions:

• Statistical models utilizing embedded measures can accurately estimate subjective TWSA.
• This approach enables cognitive state estimation without task disruption, proving valuable in operational settings.