Designing a test battery for real-world visual search

Sherman, Charli; Clarke, Alasdair; Hughes, Anna · 2025 · Crossref

DOI: 10.21203/rs.3.rs-6810586/v1

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 addresses the lack of methodological coherence and ecological validity in visual search research. While laboratory-based computer tasks have successfully refined psychological theories, there is limited evidence regarding whether these findings generalize to naturalistic, real-world scenarios involving head movement and larger search environments. To bridge this gap, the authors developed a standardized, open-source “naturalistic search task battery” using low-cost, readily available materials (LEGO bricks, bookshelves, and jigsaw puzzles). The study aims to replicate classic visual search effects in real-world settings and explore whether performance across diverse tasks correlates, indicating shared underlying cognitive processes. The researchers tested 71 participants using mobile eye-tracking technology across five tasks. Experiment 1 included three hypothesis-driven tasks: a “LEGO search” task replicating feature and conjunction search effects; a “LEGO dots” task examining set-size effects on reaction time; and a “bookshelves” task investigating the impact of scene regularity (organized vs. unorganized) and target presence. Experiment 2 featured two exploratory tasks: a “jigsaws” task assessing the influence of reference templates and search strategies, and a “LEGO building” task analyzing complex search behaviors, including planning and look-ahead fixations. Data were analyzed using Bayesian multi-level linear models, with post-hoc analyses for exploratory tasks. The results confirmed that classic visual search effects generalize to real-world contexts. In the LEGO search task, color feature searches were fastest, followed by conjunction searches, with shape feature searches being slowest, likely due to orientation variability in 3D environments. The LEGO dots task demonstrated a linear increase in reaction time with set size. The bookshelves task revealed that searches were faster in organized (homogeneous) environments and when targets were present. In the jigsaw task, participants performed better when a reference box was available, and using an “edge-first” strategy improved performance on difficult puzzles without a reference. The LEGO building task highlighted complex behaviors such as look-ahead planning and checking previous instructions, though visual search time did not significantly correlate with overall task completion time. The study concludes that a standardized battery of real-world tasks can effectively replicate laboratory findings while capturing richer behavioral data, such as strategy use and planning. Positive correlations were observed between tasks, suggesting that similar cognitive capabilities underlie diverse real-world search activities. This framework offers a reproducible, low-cost method for researchers to study visual search with higher ecological validity, facilitating broader theoretical development and comparison across studies.

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