Habit-like attentional bias is unlike goal-driven attentional bias against spatial updating

Hong, Injae; Kim, Min-Shik · 2022 · Crossref

DOI: 10.1186/s41235-022-00404-7

archive: archived pipeline: cataloged verified

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

Summary

This study investigates how the source of statistical knowledge—explicit instruction versus implicit experience—modulates the flexibility and speed of spatial attentional bias during visual search. The research addresses a fundamental question regarding the adaptability of the visual system in dynamic environments where statistical regularities change. Specifically, it tests whether habit-like attention (derived from repetitive experience) and goal-driven attention (derived from explicit top-down knowledge) are dissociable mechanisms, particularly concerning their ability to update spatial bias when target location probabilities shift. The researchers employed a location probability learning (LPL) paradigm in an online visual search task. Participants searched for a T-shaped target among L-shaped distractors. The experiment consisted of a training phase, where targets appeared frequently in one quadrant (“old-rich”), and a switching phase, where the high-probability quadrant shifted to a different location (“new-rich”). Participants were divided into two groups: an “instruction” group, which received explicit notifications about the regularity changes, and a “no-instruction” group, which had to learn the regularities implicitly through exposure. Reaction times (RTs) were analyzed using generalized linear mixed-effect models to assess attentional bias toward the old-rich, new-rich, and sparse quadrants. The results demonstrated a clear dissociation between the two attentional sources. In the training phase, both groups showed significant attentional bias toward the rich quadrant. However, during the switching phase, the instruction group rapidly readjusted their attention, showing significantly faster RTs for the new-rich quadrant compared to the old-rich quadrant. In contrast, the no-instruction group exhibited persistent bias toward the old-rich quadrant, with RTs for the old-rich and new-rich quadrants being statistically indistinguishable. This indicates that while explicit knowledge allows for flexible and rapid updating of spatial attention, implicit learning results in a more rigid, habit-like bias that is slow to discard outdated statistical information. The findings confirm that habit-like attention is relatively implicit and inflexible compared to goal-driven attention. The study refutes recent claims suggesting LPL is fully flexible and explicit, demonstrating that the source of knowledge critically determines search strategy adaptability. These results have practical implications for designing training programs in fields requiring rapid visual adaptation, such as medical imaging or cybersecurity. They suggest that explicit instruction is superior for tasks requiring quick readjustment to changing environments, whereas implicit learning may be beneficial for stable contexts but maladaptive when regularities shift rapidly.

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.

• inattentional change blindness