A comparison of human and near-optimal task management behavior

Area of Science:

• Human-Computer Interaction
• Cognitive Psychology
• Operations Research

Background:

• Previous multitasking models focused on discrete tasks.
• This study addresses continuous task management for operators.
• A novel framework uses a juggler metaphor for concurrent task performance.

Purpose of the Study:

• Develop an abstract framework for diverse multitasking scenarios.
• Introduce a quantitative model for managing multiple continuous tasks.
• Compare human multitasking performance against a normative model.

Main Methods:

• Developed the 'Tardast' low-fidelity simulator based on the juggler metaphor.
• Measured multitasking performance of 10 participants across five scenarios.
• Compared participant performance to a near-optimal solution from tabu search heuristics.

Main Results:

• The normative model (tabu search) generally outperformed human participants.
• Individual participant performance approached the normative model's level.
• Participants improved by initially managing fewer tasks effectively.

Conclusions:

• Strategic task management (selection) is more critical than tactical management (switching frequency).
• Operators may initially overreact to performance penalties.
• Findings inform the design of complex systems, equipment, and operator training.