A diffusion model analysis of target detection in near-threshold visual search

Area of Science:

• Cognitive Psychology
• Neuroscience
• Visual Perception

Background:

• Distinguishing between competing models of visual search is challenging due to similar accuracy predictions.
• Understanding how evidence is combined across locations and how evidence quality varies is crucial for visual search research.

Purpose of the Study:

• To differentiate between parallel and pooled evidence accumulation models in visual search.
• To investigate how evidence quality varies with stimulus type and contrast.
• To explore the role of attentional resource allocation in visual search.

Main Methods:

• Utilized the diffusion model to analyze response times and accuracy data.
• Conducted four near-threshold visual search tasks.
• Examined asymmetries in response-time distributions between target-present and target-absent trials.

Main Results:

• Evidence accumulation occurred in parallel for individual stimuli, not pooled into a single decision.
• Evidence quality for targets significantly improved with increasing contrast compared to non-targets.
• A fixed-capacity memory system, with preferential resource allocation to targets, explained the observed pattern.

Conclusions:

• Parallel evidence accumulation better explains visual search performance than pooled models.
• Visual search systems prioritize targets, allocating resources preferentially, especially at higher contrasts.
• Task heterogeneity can reduce the efficiency of preattentive filtering, impacting search asymmetry.