Kuniaki Mihara ran the same experiment twice, once in a room and once in a model of it. Twenty-six people sat for five minutes either with a window view or with the blinds closed, with skin temperature, skin conductance, heart rate variability and EEG recorded, and then took cognitive tests. With the view, skin temperature was 0.1 °C higher, skin conductance 0.12 microsiemens lower and heart rate 0.98 beats per minute lower, all in the direction of less stress. Short-term working memory improved by 6 per cent in the real room, significantly. In the virtual room the same improvement appeared but did not reach significance, while the physiological readings matched the real ones closely.
Sun Woo Chang and Stefano Schiavon asked what makes a view good. Sixty-nine participants each rated 15 window views inside one building, and eight attributes were tested: how many layers were visible, whether nature was present, how far the objects were from the glazing, the observer’s distance, the window-to-wall ratio, the horizontal and vertical view angles and the blind position. Views with 35 per cent or more greenery or sky and unobstructed clear glazing were strongly associated with higher satisfaction. Views without blinds were twice as satisfying as views with them. The blind position moderated everything else. Under clear glazing, adding nature and extending the visible distance both improved satisfaction; with the blinds down, neither change did anything at all.
Poletto recorded the gaze of 84 people watching 35 high-resolution window-view recordings in real time. The drivers of attention fell into a clear order. Semantic and social objects accounted for 57.3 per cent of the importance, perspective features 32.9 per cent and movement 9.8 per cent. Places that were genuinely fixated had higher energy, contrast and colourfulness than random points on the same view, by 17.1, 15.4 and 14.4 per cent.
Se Ho Park and Han Jong Jun measured openness as isovist volume and engagement as an EEG index, in 26 people moving through 16 virtual conditions. The relationship was an inverted U: engagement rose with openness to a point and fell after it, and the pattern was reproduced in 18 of the 26 participants taken individually. Park and Jun state the limits themselves: the setting was virtual reality, the sample 26 people, and the marginal R² 0.020, a very small absolute effect.
Frontczak analysed 52,980 responses from 351 office buildings collected over ten years and ranked 15 environmental parameters by how much each contributed to overall workspace satisfaction. Satisfaction with the amount of space came first, at an odds ratio of 1.57, ahead of noise level at 1.27 and visual privacy at 1.26. It ranked first for every age group, for both sexes, for single offices, shared offices and cubicles alike, and for desks within 4.6 metres of a window and beyond it. Satisfaction with the amount of space was not related to the gross amount of space available per person. More square metres per head did not raise it.
Frontczak’s 52,980 responses come from real offices; no other study here does. Chang worked with 69 people rating views inside one building, Poletto with 84 watching recordings. Mihara and Park each worked with 26, one in a room and one in virtual reality.