Statistical learning of spatiotemporal regularities dynamically guides visual attention across space

Xu, Zhenzhen; Theeuwes, Jan; Los, Sander A. · 2023 · Crossref

DOI: 10.3758/s13414-022-02573-5

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 implicit statistical learning of spatiotemporal regularities dynamically guides visual attention toward probable target locations. While prior research established that spatial and temporal regularities separately bias attention, it remained unclear whether combined spatiotemporal contingencies—where timing predicts location—guide attentional orienting or merely motor preparation. To disentangle these mechanisms, the authors employed an additional singleton task where motor responses were independent of the spatiotemporal regularities, ensuring that any performance benefits reflected attentional guidance rather than motor preparation. The research comprised three experiments using different interval distributions (uniform, exponential, and anti-exponential) to control for general temporal preparation effects. Participants searched for a unique shape among distractors and reported the orientation of a line within the target. Unbeknownst to participants, the target appeared more frequently at a specific "high-short" location after a short interval (500 ms) and at an opposite "high-long" location after a long interval (1,500 ms). The study measured reaction times (RTs) and accuracy, analyzing data using linear and generalized linear mixed models while controlling for participant awareness of the regularities. Results demonstrated that participants performed significantly better when targets appeared at high-probability locations compared to low-probability locations. Crucially, search efficiency was highest when the target appeared at the high-probability location associated with the specific interval (temporally congruent) rather than the high-probability location associated with the other interval (temporally incongruent). This benefit persisted across all three interval distributions. For instance, in Experiment 1, RTs were significantly shorter for the high-short location after the short interval compared to all other locations, indicating prioritization of the temporally congruent location. Accuracy analyses similarly showed superior performance for the high-short location. These findings held regardless of whether participants were explicitly aware of the regularities. The study concludes that implicitly learned spatiotemporal regularities dynamically guide visual attention toward probable target locations. By demonstrating this effect in a paradigm where motor responses were decoupled from the spatiotemporal contingencies, the authors provide strong evidence that the guidance is attentional rather than motoric. This extends the understanding of statistical learning, showing that the visual system integrates spatial and temporal information to optimize attentional selection in dynamic environments, independent of general temporal preparation mechanisms.

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