Collaborative search in a mock baggage screening task.

Area of Science:

• Cognitive Psychology
• Human-Computer Interaction
• Information Science

Background:

• Signal detection theory models collaborative performance in visual search.
• Real-world stimuli challenge assumptions of traditional models, like equal variance and independent observers.
• Bayesian hierarchical modeling offers a solution for complex data.

Purpose of the Study:

• To investigate team performance in a simulated baggage X-ray screening task.
• To compare team performance against predictions from information integration models.
• To explore the impact of correlated vs. independent judgments on collaborative signal detection.

Main Methods:

• Utilized Bayesian hierarchical modeling of receiver operating characteristics.
• Participants (N=28-32) worked individually and in teams of two on a simulated X-ray screening task.
• Compared team performance to two versions of a uniform weighting model (independent vs. correlated judgments).

Main Results:

• Teams underperformed predictions assuming uncorrelated judgments.
• Teams outperformed predictions based on observed correlations in individual judgments.
• Findings suggest factors beyond simple information aggregation influence team performance.

Conclusions:

• Collaborative signal detection is influenced by more than just the aggregation of individual judgments.
• Motivational factors or strategic coordination may enhance team performance in visual search.
• Future research should explore mechanisms driving improved collaborative effort and judgment decorrelation.