In 1998 Andrea Faber Taylor watched 64 outdoor spaces at an inner-city public housing estate on four separate occasions and recorded what the children were doing. Of 262 children observed, 73 per cent were playing and 87 per cent of the groups had an adult within supervising range. That is the healthy overall picture, and the paper reports it first. On the 27 spaces with little vegetation, play and access to adults were about half the level found on the 37 with more trees and grass, and creative play was significantly lower.
Yuwen Zhao asked which part of a public space does that work. In 2025, 152 Dutch adults between 18 and 33 watched videos of virtual squares in which seven features were varied systematically: people, grass, water, benches, trees, vertical greenery and traffic volume. After each video they reported how lonely, safe, relaxed and excited they felt at that moment. The presence of other people moved momentary loneliness, and so did trees and grass. The figure Zhao opens with is national statistics rather than the experiment: almost 40 per cent of Dutch people aged 15 and over reported feeling lonely in 2023, and 48 per cent of those between 15 and 25.
Anna Lane surveyed 981 people aged 55 and over across three Singapore neighbourhoods in 2018, measuring social networks on the six-item Lubben scale and setting them against the built environment as recorded from mapping data. Women living close to a wet market had larger networks, and the association held independently of age, health, education, household size and every other destination type tested. Men did not show it. A wet market is where people bargain for fresh produce. What happens at the stall is not in Lane’s data.
Won-Tae Kim tested the same room in three states. Sixty architecture students experienced an underdefined university lounge in random order: bare, with partitions added in mixed reality, and with partitions plus symbolic identity cues. Territorial comfort and control rose at each step, and strongly. Sense of community rose only in the third version, at partial eta squared between 0.17 and 0.28. Partitions on their own did nothing for it.
Laura McGrath interviewed 46 Australians during pandemic restrictions. The loss of public space showed up in four separate forms of loneliness: emotional, social, collective and existential. Some contact moved online; the unplanned encounters had no substitute in the interviews.
Faber Taylor observed children rather than asking them, which is a strength for the behaviour and a limit for the reason. Lane surveyed 981 people once and cannot show which way the arrow points. Zhao and Kim both worked in simulations, with 152 and 60 people.