ATLAS: Mapping ATtention’s Location And Size to probe five modes of serial and parallel search

Davis, Gregory · 2024 · OpenAlex-citations

DOI: 10.3758/s13414-024-02921-7

archive: archived pipeline: cataloged verified

Get this paper ↗ (DOI — opens at the source; we link to it, we don't host it)

Summary

This paper introduces ATLAS (ATtention Location And Size), a novel experimental procedure designed to directly measure the spatial location, breadth, and guidance of visual attention during search tasks. The authors argue that conventional visual search paradigms, which rely on manipulating "set size" (the number of items in a display), are flawed because they introduce stimulus confounds—such as perceptual noise and lateral interactions—that obscure the distinction between serial and parallel processing. Furthermore, standard response-time analyses often fail to differentiate between inefficient parallel processing and serial, unguided attention. To address these limitations, ATLAS aims to profile the attention "window" by tracking where attention is focused and how broadly it is distributed, allowing for the discrimination of five distinct search modes: serial-unguided, sequential-guided, unguided attention to clumps, and broad parallel attention with or without guidance. The ATLAS method employs a dual-task design where participants search through displays of six items. The search is intermittently interrupted by a brief "Memory Display" containing different stimuli (rings and circles). Participants are then asked to report the single item from the Memory Display they perceived most clearly. Crucially, their choice is restricted to a "Choice Set," indicated by green numbers on a subsequent display. By varying the size and spatial arrangement of this Choice Set across conditions, the researchers can infer the distribution of attention during the search phase. Three primary conditions were used to assess attention breadth: "1ofN" (one reportable item, serving as a baseline for average attention), "NofN" (all items reportable, allowing participants to choose the peak of attention), and "1of1" (one item presented, testing for exogenous cueing effects). Supplementary conditions, such as "3ofN," varied the spatial arrangement of reportable items to refine estimates of attention breadth. The study found that ATLAS effectively distinguishes between narrow (serial) and broad (parallel) attention. Under narrow attention, performance in the NofN condition was significantly higher than in the 1ofN and 1of1 conditions, creating a "downward-elbow" pattern in accuracy scores. This indicates that when attention is focused on a single item or small clump, participants can only report items at that specific location, resulting in poor performance when the reportable item is elsewhere. Conversely, under broad attention, the expected "upward-elbow" pattern did not fully materialize; instead, 1of1 accuracy fell between NofN and 1ofN performance. While this underestimated the peak performance relative to predictions, the distinct patterns across conditions successfully identified the predominant mode of search. The method also tracked attention guidance, signaling when attention was directed toward targets via local or global parallel processing. The significance of this work lies in providing a direct behavioral measure of attention dynamics that circumvents the confounds of set-size manipulations. By profiling the attention window, ATLAS offers a tool to differentiate between serial and parallel processing mechanisms more precisely than response-time slopes alone. This approach supports more sophisticated modeling of visual search, including hybrid mechanisms like Wolfe’s Guided Search. Although the initial investigation used highly regular stimuli, the authors conclude that ATLAS has broader potential for investigating attention guidance and breadth in complex search scenarios, offering a complementary measure to traditional psychophysical methods.

Provenance

The full processing record for this entry. Every stage of this paper's journey through the pipeline is logged — what ran, with which tool and model, how many attempts it took, and when it last completed.

Summary generated by qwen3.6-27b-prismaquant on 2026-06-25; verification: verified.

Topics

Ranked by relevance to this paper. Hover a topic for its definition.

• visual search
• inattentional change blindness
• useful field of view
• attention selective divided