Spatial perception in architectural transitions: A psychometric study of the compression-expansion effect in ceiling-height
Year: 2026
Type: article
Abstract: We constantly move between rooms, transitioning between spaces with different spatial features. When entering a high-ceilinged space through a lower one, we experience the compression-expansion effect, a well-established architectural principle long used to amplify perceived spaciousness. Despite its prominence in architectural theory and practice, the underlying perceptual psychometrics remain poorly understood. Here we identify and specify the psychometric function governing how sequential changes in ceiling height shape spatial perception. Using immersive virtual reality with freely moving participants, we manipulated room-height ratios and asked participants to adjust perceived room height following exposure to a preceding space. Perception was moderately better described by a nonlinear sigmoid model than by a linear relationship between ΔHeight and ratio, suggesting a saturating, asymmetric pattern: compression was overestimated by 31.44% (nominal ratio 0.8), whereas expansion was overestimated by 11.68% (nominal ratio 1.2). These dynamics align with predictive accounts of perception and parallel serial dependence effects in vision, indicating that architectural experience is temporally extended and shaped by prior spatial context. We provide the first quantitative characterization of the compression-expansion effect, transforming a long-standing design principle into an evidence-based model with implications for human experience, material efficiency, and sustainable architectural practice.
Authors: Aja Khalid Hussein, Dylan Chau Huynh, Amir Kalantarifard, Farzaneh Behzadi, Xinyi Meng, Stine Louring, Lars Brorson Fich, Nicolai Carlberg, Benedicte Mygind, Gitte Frederiksen Kjølby, My Lunsjö, Tandia Hardcastle, Nicolaj Østergaard Thunbo, Camilla Jensen Thorup, Cecilie Marie Busk, Katarina Bramsen Buhl, Klaudio Muca, Stine Lea Jacobi, Anne Gade Iversen, Zakaria Djebbara
Language: English
Subfield: Cognitive Neuroscience
Field: Neuroscience
Domain: Life Sciences
Open Access status: hybrid
Estimated APC paid: $4,380
Funders: Aalborg Universitet, Realdania
OpenAlex ID: W7130720852