Attentional suppression is in place before display onset

Huang, Changrun; Donk, Mieke; Theeuwes, Jan · 2023 · Crossref

DOI: 10.3758/s13414-023-02704-6

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 the temporal dynamics of attentional suppression induced by statistical learning, specifically determining whether suppression of a high-probability distractor location is implemented proactively before display onset or reactively at the moment of expected onset. While previous research established that observers learn to suppress locations frequently containing salient distractors, it remained unclear if this suppression is continuously active or flexibly activated only when the display is anticipated. The authors aimed to distinguish between these possibilities by probing attentional distribution at varying time points relative to the expected search display. The experimental design involved 75 participants performing a visual search task combined with an occasional probe offset detection task. In the search task, participants identified a target shape singleton amidst distractors, with the distractor appearing more frequently (65%) at a high-probability location (HPL) to induce statistical learning. Crucially, the search display was consistently presented 800 ms after a placeholder display. In the probe trials, the placeholder duration was varied (400, 800, or 1,200 ms) to present the probe display before, simultaneously with, or after the expected search onset. Participants detected probe offsets to measure attentional allocation at these specific time points. Statistical analysis utilized linear mixed models to assess reaction times and accuracy, controlling for factors such as target location and participant awareness of the distractor regularity. The results confirmed that statistical learning led to effective suppression of the HPL, evidenced by faster and more accurate search performance when the distractor appeared at this location. Furthermore, target search was slower when the target appeared at the HPL, indicating feature-blind spatial suppression. The critical finding from the probe task was that probe detection was consistently slower at the HPL compared to low-probability locations, regardless of the stimulus onset asynchrony (SOA). There was no significant interaction between probe location and SOA, demonstrating that suppression was already in place 400 ms before the expected display onset and persisted 400 ms after. This effect held true irrespective of whether participants were consciously aware of the distractor regularity. The authors conclude that statistically learned suppression is proactively implemented well before the availability of the search display, affecting the weights within the spatial priority map early in time. This finding challenges the notion that suppression is flexibly activated only at the moment of expected onset, as suggested by some models of top-down template activation. Instead, the results support a model where statistical learning continuously biases the priority map, deprioritizing high-probability distractor locations prior to display onset. This proactive suppression operates independently of explicit awareness and distinct from feature-based signal suppression, highlighting the robust and early nature of history-driven attentional biases in visual selection.

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