Designing a test battery for real-world visual search.

Area of Science:

• Cognitive Psychology
• Human-Computer Interaction
• Computer Vision

Background:

• Laboratory-based visual search studies offer limited insight into real-world search behavior.
• Existing research on naturalistic search is fragmented across disciplines, hindering data comparability.

Purpose of the Study:

• To develop a standardized, open-source 'naturalistic search task battery' for reproducible research.
• To enable cross-study data generation and advance theory/modeling of real-world visual search.

Main Methods:

• Development of a suite of standardized real-world search tasks (Lego, bookcase, jigsaw, Lego building).
• Validation of tasks by replicating known findings (set-size effects) and demonstrating new ones (scene regularity).

Main Results:

• Replicated set-size effects and search condition differences using Lego tasks.
• Demonstrated effects of scene regularity in a bookcase search task.
• Explored strategic effects in complex jigsaw and Lego building tasks.

Conclusions:

• The naturalistic search task battery provides a foundation for standardized research on real-world visual search.
• The tasks allow for the investigation of cognitive strategies and the generalizability of findings across diverse search scenarios.
• Future work can explore the predictive power of performance across tasks to understand underlying cognitive capabilities.