One factor to bind them all: visual foraging organization to predict patch leaving behavior with ROC curves.

Area of Science:

• Cognitive Psychology
• Human-Computer Interaction
• Visual Perception

Background:

• Determining when to stop searching is crucial in visual tasks like medical screening or security screening.
• Current models often rely on intake rates, which may not fully capture optimal stopping behavior in complex visual search.

Purpose of the Study:

• To evaluate the predictive power of search organization indexes (e.g., best-r, mean ITD, PAO) for optimal quitting behavior in visual search.
• To compare the effectiveness of organization measures against traditional intake rates.
• To investigate the combined predictive power of organization and intake rates.

Main Methods:

• Conducted experiments comparing static and dynamic foraging in 29 adults.
• Reanalyzed existing data from diverse lifespan foraging tasks.
• Utilized Receiver Operating Characteristic (ROC) curves to analyze predictive accuracy.

Main Results:

• Search organization indexes consistently outperformed classic intake rates in predicting human quitting behavior.
• A combined factor incorporating both organization and traditional intake rates proved to be the most effective predictor.
• Results were replicated across diverse datasets and foraging tasks.

Conclusions:

• Search organization measures are superior to traditional intake rates for predicting optimal stopping in visual search.
• A hybrid approach combining search organization and intake rates offers the best predictive model.
• This research opens new avenues for understanding decision-making in visual search based on search organization.