The Hidden Geometry of Attention: Viewing Distance Is Associated with Flanker Interference
Abstract
Visual-attention tasks are often specified in degrees of visual angle, yet viewing distance may vary during testing, altering the realised angular geometry of otherwise identical displays. We tested whether trial-level viewing distance was associated with performance in the Attention Network Test–Revised, focusing on reaction-time flanker conflict. Eighty-six adults were recruited; the computer-based analyses included 59 participants with valid camera-derived viewing-distance data, contributing 12,059 trials, of which 11,414 correct trials met the reaction-time criteria. Mixed-effects models separated within-participant from between-participant variation in viewing distance. Realised target eccentricity varied across trials (median = 3.27°, interquartile range = 2.96°–3.73°). Flanker conflict was larger on trials when participants were farther from the display than their own mean viewing distance (flanker congruency by within-participant distance: b = 0.022 per 10 cm on the log-reaction-time scale, Benjamini–Hochberg-adjusted p = .007). The corresponding effective-angle model showed greater flanker conflict at smaller realised target eccentricities. Within-participant viewing distance did not significantly moderate the reaction-time spatial cue-validity effect, and no corresponding flanker-by-distance interaction was detected for response accuracy. An exploratory virtual-reality (VR) task analysis in 79 participants showed a corresponding pattern, with greater flanker conflict at farther-than-usual viewing distances and smaller realised target eccentricities. Thus, viewing geometry was associated with reaction-time flanker conflict, with convergent evidence from an exploratory VR extension. Measuring or constraining viewing distance may improve the interpretation and reproducibility of visually presented cognitive tasks when stimulus spacing and eccentricity are theoretically relevant.
Related
- Differential effects of free viewing and goal-directed search on visual attention: An eye-tracking study
- Neural tracking of surprisal and semantic distance in naturalistic movie viewing
- Viewing geometry and pattern type shape the detectability of snakes
- Similarity in greater Chinese ideology is associated with intersubject neural similarity during viewing of a pro-China video
- Transspinal temporal interference stimulation enhances neuroplasticity-associated responses and facilitates motor recovery after spinal cord injury in mice
BibTeX
@misc{tekampe2026hidden,
title = {{The Hidden Geometry of Attention: Viewing Distance Is Associated with Flanker Interference}},
author = {David Levi Tekampe and Philip Samuel Santangelo},
howpublished = {PsyArXiv},
year = {2026},
month = sep,
doi = {10.31234/osf.io/bsn37_v3},
url = {https://osf.io/preprints/psyarxiv/bsn37_v3/},
note = {CC BY 4.0},
}
RIS
TY - UNPB
TI - The Hidden Geometry of Attention: Viewing Distance Is Associated with Flanker Interference
AU - David Levi Tekampe
AU - Philip Samuel Santangelo
PY - 2026
DA - 2026/09/07
PB - PsyArXiv
DO - 10.31234/osf.io/bsn37_v3
UR - https://osf.io/preprints/psyarxiv/bsn37_v3/
AB - Visual-attention tasks are often specified in degrees of visual angle, yet viewing distance may vary during testing, altering the realised angular geometry of otherwise identical displays. We tested whether trial-level viewing distance was associated with performance in the Attention Network Test–Revised, focusing on reaction-time flanker conflict. Eighty-six adults were recruited; the computer-based analyses included 59 participants with valid camera-derived viewing-distance data, contributing 12,059 trials, of which 11,414 correct trials met the reaction-time criteria. Mixed-effects models separated within-participant from between-participant variation in viewing distance. Realised target eccentricity varied across trials (median = 3.27°, interquartile range = 2.96°–3.73°). Flanker conflict was larger on trials when participants were farther from the display than their own mean viewing distance (flanker congruency by within-participant distance: b = 0.022 per 10 cm on the log-reaction-time scale, Benjamini–Hochberg-adjusted p = .007). The corresponding effective-angle model showed greater flanker conflict at smaller realised target eccentricities. Within-participant viewing distance did not significantly moderate the reaction-time spatial cue-validity effect, and no corresponding flanker-by-distance interaction was detected for response accuracy. An exploratory virtual-reality (VR) task analysis in 79 participants showed a corresponding pattern, with greater flanker conflict at farther-than-usual viewing distances and smaller realised target eccentricities. Thus, viewing geometry was associated with reaction-time flanker conflict, with convergent evidence from an exploratory VR extension. Measuring or constraining viewing distance may improve the interpretation and reproducibility of visually presented cognitive tasks when stimulus spacing and eccentricity are theoretically relevant.
ER -