This evidence sheet is published by The Built Review.
The psychologist Nooshin Momenzadeh ran four experiments on whether liking a room helps you remember it, varying ceiling height, openness and curvature. In a laboratory version, 40 people chose like or dislike for interior images. High ceilings were preferred 64% of the time against 48 for low ones, and rectilinear layouts 67% against 45 for curvilinear. In the last experiment, fifty people were recruited on Prolific and 48 completed both blocks, rating 60 interior images and then, eight days later, picking them out of 120. Preference predicted recognition accuracy only in this version with the delay; in the three earlier experiments, where recall was tested immediately, it did not, and the paper attributes that to a ceiling effect in performance rather than to an absence of the phenomenon. The preference effect was weaker for low ceilings and closed spaces.
The architecture researcher Zhihui Zhang asked 183 architecture students at the Universitat Politècnica de Catalunya to work with seven ceiling heights, from 2.26 to 9.57 metres, in virtual art galleries. In the first study each group rated the emotions a single assigned height produced. In the second, 178 of them were given an emotion and asked to choose the height that fitted it. For anger, 44.44% chose 2.26 metres, the lowest on offer, and for fear 37.93% chose the same. For joy, 34.48% chose 4.79 metres. Lower ceilings drew fear and anger, higher ceilings joy, and the remaining emotions varied without a clean pattern. The heights come from the Modulor series and are taken from the Joan Miró Foundation and the National Museum of Western Art in Tokyo, so they are gallery dimensions rather than office ones.
The design researcher Alp Tural put eye trackers on 35 students assessing 16 photorealistic interiors, once on a screen and once in virtual reality. Walls took 50.1 to 56.7% of the normalised gaze duration in every condition. Ceilings came second, between 16.6 and 25.8%, with the angled ceiling drawing the most attention and the curved one the least, flat at 17.2 and vaulted at 21.6. View access and materiality both significantly changed how spacious the room felt. View content and ceiling geometry did not reach significance, so the ceiling that drew the most gaze changed nothing about the sense of space.
In a companion study on the same 16 interiors and the same 35 students, ceiling geometry did matter for a different outcome. Varied ceilings, and vaulted ceilings in particular, were rated more restorative than flat ones, and the difference was larger in virtual reality than on the screen. Larger windows raised restorativeness in both modes, views of water rated most restorative and views of woodland second, and natural materials beat other materials, especially in virtual reality.
The architecture researcher Anbang Dai shortened the looking time instead. Participants were shown images of interiors for 200 milliseconds, long enough for a single fixation, and asked for an aesthetic judgment. Ceiling height, openness and contour influenced that judgment as clearly at 200 milliseconds as they did at 3,000, which is the standard exposure in this field. The version of the paper available here does not report how many people took part.
Every study here worked with images or virtual rooms and none with a room people use. Zhang's 183 architecture students were rating galleries. Momenzadeh's preference figures come from 40 people in a laboratory, and both Tural studies from the same 35 students. Dai's paper does not report how many took part.
