To quit or not to quit in dynamic search

Area of Science:

• Cognitive psychology
• Visual perception
• Human-computer interaction

Background:

• Visual search efficiency is typically measured by reaction-time (RT)/set-size slope.
• Dynamic visual search, where items reshuffle, has unclear underlying mechanisms despite similar target-present slopes to static search.
• Previous research overlooked the flat target-absent slope in dynamic search.

Purpose of the Study:

• To investigate the mechanisms behind dynamic visual search performance.
• To explain the differential reaction-time (RT)/set-size slope patterns in dynamic versus static search.
• To develop and validate a computational model for dynamic search.

Main Methods:

• Conducted three experiments comparing dynamic and static visual search displays.
• Manipulated factors including reward for correct responses and search difficulty.
• Analyzed search sensitivity and response criteria to inform model development.

Main Results:

• Confirmed a target-absent:target-present slope ratio near or below 1 in dynamic search, versus above 2 in static search.
• Observed consistent slope ratios across reward conditions and difficulty levels.
• Developed a multiple-decisions model that accurately predicts these differential slope patterns.

Conclusions:

• The "quit-the-search" decision, based on target absence likelihood, is crucial for the 1:1 slope ratio in dynamic search.
• Stopping thresholds are linearly related to set size and reward, further explaining dynamic search dynamics.
• The multiple-decisions model provides a robust framework for understanding visual search in dynamic environments.