Expertise Scientific research on how buildings
Scientific research on how buildings
affect human health and performance.
The argument
Almost none of it reaches the people making building decisions at the point where the decision is still open. That is not a knowledge problem. It is a timing problem, and it is the one we work on.
We did not invent evidence grading.
We use GRADE. It is the international standard for assessing the quality of evidence, it is what the WHO and Cochrane use, and we had no part in developing it. HIGH means strong study design, adequate sample, consistent results. MODERATE, LOW and VERY LOW are labelled as such and used accordingly.
We say this first because it is the obvious question, and because a practice that grades evidence for a living should not make ungraded claims about itself.
There are many evidence systems. Ours is the one you can check.
Every finding we use is published at thebuiltreview.com with its study, its DOI, its grade and its contradicting studies.
Every finding we rely on is readable before you speak to us. You can take our own material and argue against us with it. We think that is the only honest way to make a claim about evidence, and it is the reason the publication is kept separate from the practice.
One number, for organisations whose largest cost is people.
Air quality alone takes 5 to 10% off performance, under conditions today counted as acceptable. Doubling the outdoor air supply costs the building owner between 8 and 40 dollars per person per year.
Wargocki & Wyon, 2017, Building and Environment · MacNaughton et al., 2015, IJERPH
For most institutions the salary bill is a multiple of the rent bill. A change in absence and performance is therefore not measured against the cost of the building. It is measured against the cost of everyone in it.
Set against that, 5 to 10% is not a wellbeing outcome. It is an operating result.
Selected findings
A selection, grouped by what it affects. The evidence we produce for a client is specific to their project, their building type and their users.
Cognition
101%
Higher cognitive scores in a green building with high outdoor air ventilation
Allen and colleagues tested the same people over several days; the green building alone accounted for 61%.
12%
Faster information processing when CO₂ falls from 2100 to 900 ppm
A second study inside the same finding: a synthesis of classroom data, by different authors and with a different design, pointing the same way.
+42%
Higher scores on decision-making tests for office workers in the better-daylit of two identical offices
The same participants slept 37 minutes longer a night.
−7%
Lower learning performance at 33°C
Temperature acts non-linearly, and both ends of the comfortable band cost performance.
Health and absence
16 → 6
Work hours a month lost to asthma and allergies, before and after a move into a healthier building
The samples reporting were 46 and 33 people, so attrition is part of the picture.
0.64
Odds ratio for wheezing after damp and mould are remediated
A Cochrane review of remediation trials.
0.5–0.9%
Fewer illness symptoms per 1000 ppm of CO₂ removed
Measured across school and office buildings.
Recovery
d −0.7
Lower anxiety with a view of nature, up to d −2.42
Consistent across eleven studies, from hospital wards to workplaces.
7.3 h
Shorter hospital stay for every extra 100 lux of daylight in the room
Measured with sensors in the rooms of 263 bypass patients.
Behaviour
×2
Self-reported work performance lost to noise, after moving to open plan
Questionnaire data from one organisation before and after a relocation.
d −0.7
Lower working memory under background noise, up to d −2.65
Strongest in children and in anyone working in a second language.
1.57
Odds ratio for satisfaction, driven by how much space a person has
Ahead of visual privacy at 1.26 in the same study.
A selection of 55 sheets, each on a single question.
Every finding names its author, the journal, the year and the DOI. The full library is larger.
Air, noise and a day's output 23% of 7,441 workers in 167 European office buildings rated their overall comfort below four out of seven. Of 7,441 workers in 167 modern European office buildings, 23 per cent rated their overall comfort below four out of seven, and noise was the commonest reason. 5 studies Attachment to a place 60% of the variance in sense of community was explained by age, income and how restorative people found the place. Five studies on what makes people feel they belong where they live. The largest has 35,005 respondents and reports no single figure. 5 studies Autistic people and the buildings they use −1.71 vs −0.85 average thermal sensation votes for autistic participants and a control group under the same slightly cold condition. Under the same cold condition, autistic participants rated the room at −1.71 on the thermal sensation scale and the control group at −0.85. Under the warm one it was 1.53 against 0.90. 6 studies Being visible to other people 58% lower odds of a good rating for working productively, for each half standard deviation of desks in the line of sight above the average of 66. The average desk in one headquarters had 66 other desks in its line of sight. Half a standard deviation more cut the odds of a good rating for working productively by 58 per cent. 5 studies Ceiling height 64% of ratings favoured high ceilings, across 60 interior images rated online by 48 people. Low ceilings took 48 per cent of the ratings. Tural tracked eye movements and found the ceiling changed where people looked but not how spacious the room felt. 5 studies Daylight at work 5–6% rent premium for Manhattan office space with high daylight against space with low daylight. Seven studies on the light itself rather than the view, one of them measuring what 170 indoor workers get in winter. 7 studies Deciding in bad air 1,000 ppm of carbon dioxide was enough to push six of nine decision-making scales down moderately, against 600 ppm. Twenty-two people made worse decisions at 1,000 ppm of carbon dioxide. Fifteen people showed nothing at all below 5,000. 5 studies Feeling safe enough to go out 50% vs 17% of women and of men aged 50 and over reported a fear of falling outdoors, 43 per cent overall. Five studies on what makes a street feel safe, including one that ranked cleaning and planting against giving an alley something to do. 5 studies Finding your way through a building 153.82 s to reach the nearer staircase from a floor plan that was hard to read, longer than others took to reach the farther one from a legible plan. With a floor plan that was hard to read, people took 153.82 seconds to reach the nearer staircase, longer than others took to reach the farther one from a plan they could read. 5 studies Getting outside during the working day 40 min more standing per day during office hours, seven months after a move into a biophilic office. Six studies on nature at work, one of them with 9,386 people and the rest with between 8 and 94. 6 studies Green around the workplace 14.6 vs 13.5 min to exhaustion on a cycling test after 90 minutes in a forest and after 90 minutes in an industrial environment. Five studies on the greenery outside the door, from a meta-analysis of city-scale data down to 25 people on a cycling test. 5 studies Green space and loneliness 29.4% of 3,043 Australians who were lonely at the start were no longer lonely 16 months later. Four studies with loneliness itself as the outcome. The largest followed the same people for sixteen months: time in nature helped them out of it without stopping anyone falling in. 4 studies Green walls and living facades 1/3 the speed at which negative mood built up during desk work with a living wall in the room, against a plain office. Five studies on plants built into the fabric of a building, four of which measure the building rather than the person. 5 studies Greenery in the room 6.5 mmHg lower diastolic blood pressure after 15 minutes repotting a plant than after 15 minutes of computer work. Six studies on plants indoors, from 24 men repotting a Peperomia to two call centres where the effect did not appear. 6 studies Growing old in the neighbourhood 21% lower risk of social isolation four years after joining Yokohama's citywide walking-points programme, among 3,171 residents aged 65 and over. Five studies on ageing in place, running from a cohort of 3,171 in Yokohama down to eleven people walking two routes. 5 studies How closed a room is 47% of occupants were satisfied with visual privacy across three activity-based offices, the lowest rating of any feature. Visual privacy was the worst-rated feature of three activity-based offices, at 47 per cent satisfied, below storage and below acoustic privacy. 5 studies How crowded a place feels 3.7–18.3% higher reported loneliness and isolation among private renters in shared housing than among other household types. In two Johannesburg neighbourhoods 57.6 per cent of dwellings were overcrowded in 2006, and the same share in 2016. 5 studies How far speech reaches 15 m → 7 m the radius of distraction before and after a pink-noise masking system at 44 dBA was switched on in an open-plan department. A masking sound of 44 dBA cut the radius over which speech distracts from 15 metres to 7, in a department of fifteen people. 5 studies How full an office is 20% of 1,403 office workers across 14 open-plan offices were dissatisfied with the physical environment, certified building or not. The 42 per cent utilisation figure comes from Strategy Plus, a consulting and design firm, by way of the introduction to a systematic review. No study in this literature measured it. 5 studies How much nature is enough 25% window-to-wall ratio is enough for a direct view of a good scene, against roughly 40 per cent for a poor one. Four studies on the dose, and each of the four measures it in a different unit. 4 studies How private a workplace feels 95% of 175 employees in 17 open-plan banks were annoyed by background noise. Of 175 employees in seventeen open-plan banks, 95 per cent were annoyed by the background noise, and the speech in the room was more intelligible than the standard calls satisfactory. 5 studies Intelligible speech and concentration 21.9% better performance on a concentration task when employees moved from the active zone of an office into an individual room. Moving from the noisy zone of an office into an individual room raised concentration-task scores by 21.9 per cent. Two of the studies here found no effect of office noise on performance at all. 5 studies Living alone OR 3.08 for loneliness among people living alone against people living with others, pooled across four cohorts. Seven studies on the household of one, including one in which every environmental risk to falling was significant for people living alone and for nobody else. 7 studies Loneliness measured where people live 27.44% of 8,638 children had passing loneliness across four years, and 2.43 per cent the lasting kind. Five studies that put loneliness next to the home itself, and one of them found the building explains the passing kind and not the lasting kind. 5 studies Measuring speech privacy 25% more strongly than the noise itself, having nowhere to speak unheard predicted acoustic dissatisfaction across 28 offices and 349 people. Across 28 offices and 349 people, having nowhere to speak unheard predicted acoustic dissatisfaction 25 per cent more strongly than the noise itself did. 5 studies Men and women in the same room −0.61 vs 0.23 the thermal sensation at which men and women were most productive in the same office. Men in one office were most productive at a thermal sensation of −0.61, on the cool side of neutral. Women in the same office were most productive at +0.23. 5 studies Moving from a cell into open plan 2× the self-rated loss of work performance to noise after 31 workers moved from private rooms into open-plan offices. Thirty-one people moved out of private rooms into an open-plan office and their self-rated loss of performance to noise doubled. The cooperation the move was for became less pleasant. 5 studies Moving into an activity-based office 36.4% of staff reported regularly using the height adjustment seven months after moving into an office designed for movement, down from 45.5 per cent at three months. Seven months after a move into an office designed for standing, 36.4 per cent of the staff were still using the height adjustment on their desks. 6 studies Nature sound in the office 86% of 198 people called a virtual office pleasant with natural sound in it, against 42 per cent with office sound alone. Four studies on birdsong and running water at work, all four run in a booth or in virtual reality. 4 studies Nature where there is no window 22.2% higher focus in an office with an artificial window than in one with no window, measured as heart rate variability. Six studies on rooms with no window, from 195 pictures counted on office walls in 1986 to virtual reality at the South Pole. 6 studies Older people and the same building 30% lower scores on Chinese speech intelligibility for older listeners than for younger ones, under the same acoustic conditions. Under identical acoustic conditions, older listeners scored 30 per cent lower on Chinese speech intelligibility than younger ones. 6 studies Open plan and cellular offices −0.86 units of workspace satisfaction across 303 buildings, where open-plan occupants gain 0.21 from easier interaction and lose 1.07 to acoustics and privacy. Across 42,764 responses from 303 office buildings, open-plan occupants gain 0.21 units of workspace satisfaction from easier interaction and lose 1.07 to acoustics and privacy. 5 studies Openness and the view out 6% better short-term working memory with a window view than with closed blinds, in a real room. A window view instead of closed blinds improved short-term working memory by 6 per cent in a real room. In a virtual copy of the same room the improvement was the same size and did not reach significance. 5 studies Personal control over the workspace 26 °C was the temperature at which task speed peaked, among 56 people tested at 23, 26 and 29 °C with a fan they controlled themselves. With a fan they could control, 56 people in Singapore were as comfortable at 26 and 29 °C as at the usual set point of 23. 5 studies Personality and the room −13% in the probability of a bad mood for every 5 °C more maximum temperature among people scoring high on neuroticism, against a 7 per cent drop overall. Every 5 °C of maximum temperature cut the probability of a bad mood for the whole day by 7 per cent overall and by 13 per cent among people scoring high on neuroticism. Among people with schizophrenia it went the other way. 5 studies Satisfaction in the office 81% of the people in over 600 office buildings were dissatisfied with at least one aspect of their workspace, acoustics most often. Eighty-one per cent of occupants of over 600 office buildings were dissatisfied with at least one thing, and acoustics led the list. 5 studies Sensitivity to noise and light 25% of 1,934 Italian remote workers named the noise other people make as the main source of disturbance. A quarter of 1,934 people working from home named the noise other people make as their main disturbance. Fifty-seven per cent described themselves as sensitive to noise. 5 studies Sick days and leaving early HR 1.95 for disability retirement among Norwegian open-plan office workers against workers in cellular offices, and 1.52 for shared offices. Open-plan office workers in Norway retired on disability at 1.95 times the rate of people in cellular offices. In one Platinum-certified building, sick leave did not fall at all. 5 studies Sit-stand desks and prompts 82 min less sitting per day in total, leisure included, with a sit-stand desk. At work alone it was 100 minutes. A Cochrane review of 34 trials puts sit-stand desks at 100 minutes less sitting a workday, and rates its own evidence low quality. 6 studies Steps in a working day 1,029.99 more steps per workday from software that prompts people to get up, measured by device across 18 trials. 614 German desk workers walk for 12.9 per cent of their working hours and would like to walk for 22.8. 5 studies Taking a break 10 min of sitting or walking somewhere green moved psychological and physiological measures of wellbeing across 14 studies, 13 of which used college students. Ten minutes of sitting or walking somewhere green was enough to move the measures in fourteen studies, and thirteen of the fourteen used college students. 5 studies Taking the stairs 52% higher odds of using the stairs after a motivational sign, pooled across 50 studies. In absolute terms the median rise was 2.2 percentage points. Fifty sign-at-the-stairs studies since 1980 move stair use by a median 2.2 percentage points, from a base of 8.1 per cent. Bauman asks researchers to stop running them. 5 studies Temperature, air and performance R² = 0.05 when the standard temperature-performance curve is tested against 35 studies of office work between 18 and 34 °C. Across 35 studies, Jose Ali Porras-Salazar could find no relationship between air temperature and office work anywhere between 18 and 34 °C. 6 studies The public space people stay in about half as much play and half as much contact with adults on the estate's 27 barren outdoor spaces as on the 37 with trees and grass. Five studies on outdoor space and who stays in it, the oldest of them a count of 262 children at play in 1998. 5 studies The same room in different countries 21.2–24.8 °C the band in which 80 per cent of occupants of European classrooms were comfortable, against 16.0 to 21.6 °C in Darjeeling. Eighty per cent of people in European classrooms were comfortable between 21.2 and 24.8 °C. In Darjeeling the same share was comfortable between 16.0 and 21.6. 5 studies The shape of a room d = 0.82 for positive affect in a curved virtual room against a rectangular one, across 35 people. In a replication with 42 people and better-matched rooms, two of 33 measures differed, both favouring the angular room. Thirty-five people felt better in a curved virtual room than a rectangular one. Tawil rebuilt the test with 42 people and better-matched rooms, and of 33 measures only two differed, both favouring the angular room. 5 studies The view from the window 3× as high a satisfaction rating for views with a lot of nature in them as for views with little, across 15 windows in one building. Five studies on what a window shows, from 15 windows rated inside one building to a climate chamber held at 28 degrees. 5 studies Ventilation and sick days OR 0.62 for adults reporting a day off work after their house was insulated, in a blinded trial of 1,350 households. Insulating 1,350 New Zealand houses raised bedroom temperatures by half a degree in winter, and adults in them took fewer days off work at an odds ratio of 0.62. 5 studies Water in the built environment 2.5× as high a rating of the acoustic environment when people could see the water they could hear as when they could only hear it. Five studies on water indoors and out, from 60 colour slides in 1981 to 7,975 smartphone assessments in 2022. 5 studies What pushes people out of the office 923 employees at four Finnish workplaces, where too few rooms for quiet work and too little storage went with wanting to telework more. Rooms for spontaneous collaboration were short as well. Wellbeing and work ability did not predict the preference at all. 5 studies When mistakes happen 26.3% higher probability of a bad move for every 10 µg/m³ more fine particulate matter indoors. Every 10 micrograms per cubic metre of fine particulate in a tournament hall raises the chance of a bad chess move by 26.3 per cent, across 30,000 moves scored by an engine. 5 studies Which place supports the work 57,286 employees surveyed across 2020 and 2021, whose choice of where to work followed how well a place supported the specific task. Across 57,286 employees surveyed between March 2020 and May 2021, support for concentrated work and support for collaborative work both related to perceived productivity, at home and at the office alike. 5 studies Wood in the room 5–16 s faster completion on all five neurobehavioural tests in wooden rooms than in a concrete one. Five studies on timber and other natural materials indoors, three of which ask people to rate their own productivity. 5 studies Working from home during the pandemic 56% higher social isolation scores for home workers than for those working on site, and 17 per cent higher for hybrid workers. Nineteen survey rounds, 41,147 Dutch workers, and home workers 56 per cent higher on social isolation than people working at a location. 4 studies Working near each other about 70% less time spent interacting face to face after two corporate headquarters moved to open plan, measured by badge. Two headquarters took the walls out to get people talking, and face-to-face contact fell by about 70 per cent. 5 studies
View the full library of 18,180 studysheets, 517 factorsheets and 67 worksheets on The Built Review → We start with a Decision Brief.
We take three weeks and charge a fixed fee. The report is yours even if we stop there.
What that involves →