Attentional capture by real and illusory faces: a failure to replicate

Miti, Francesca; Ciaramidaro, Angela; Rubichi, Sandro; Iani, Cristina · 2025 · Crossref

DOI: 10.1007/s00426-025-02211-3

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 whether the automatic attentional capture typically associated with real faces extends to illusory faces (face pareidolia) and whether this capture is truly mandatory or influenced by task demands. Prior research suggests that faces are prioritized in visual processing due to their social relevance, often capturing attention even when task-irrelevant. This prioritization is thought to rely on a broadly tuned mechanism sensitive to minimal face-like configurations, such as the T-shaped arrangement of eyes and mouth. Consequently, illusory faces—objects perceived as faces—were hypothesized to similarly capture attention automatically. The authors aimed to test this hypothesis and replicate previous findings of attentional capture by irrelevant faces, specifically examining if a minimal face configuration alone is sufficient to disrupt visual search performance. The researchers conducted three behavioral experiments using a visual search paradigm. In Experiments 1A (online, N=36) and 1B (lab-based, N=30), participants searched for a butterfly target among a circular array of six distractors. The distractors included either a real face, an illusory face, or standard objects. The goal was to determine if the presence of these task-irrelevant faces slowed reaction times compared to standard object distractors. Experiment 2 (N=30) modified the design by making the faces, illusory faces, or butterflies the targets themselves, presented in different blocks, to assess detection efficiency directly. Stimuli were carefully selected and validated; illusory faces were chosen based on high ratings of face-likeness and happy expressions to match the real face stimuli. Participants also completed the Glasgow Face Matching Task to control for individual differences in face perception abilities. The results contradicted the hypothesis of automatic attentional capture. In Experiments 1A and 1B, neither real nor illusory faces significantly slowed reaction times or increased error rates compared to standard object distractors. Bayesian analyses provided moderate to strong evidence against the inclusion of distractor type or its interaction with trial type in the models. Post-task questionnaires revealed that while most participants noticed real faces, fewer noticed illusory faces, yet this awareness did not correlate with performance interference. In Experiment 2, real faces were detected more efficiently than illusory faces or butterflies. Crucially, illusory faces showed a search disadvantage, being detected more slowly and less accurately than butterflies. These findings indicate that the mere presence of a face-like configuration does not automatically capture attention when the stimulus is task-irrelevant. The significance of these findings lies in challenging the assumption that attentional capture by faces is an entirely automatic, domain-specific mechanism driven solely by low-level visual features. The failure to replicate previous studies showing interference from irrelevant faces suggests that such capture may be contingent on contextual factors or task demands rather than being mandatory. Furthermore, the results demonstrate that illusory faces do not share the same privileged access to attention as real faces; instead, they may even hinder detection compared to non-face objects. This implies that the neural mechanisms underlying face detection are more complex than previously thought, requiring more than just a T-shaped configuration to trigger prioritized processing, and that the "face effect" in attention is not as robust or automatic as earlier literature suggested.

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 2026-06-26; verification: verified.

Topics

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

• inattentional change blindness