Usha Satish put 22 people through three two-and-a-half-hour sessions in an office-like chamber at 600, 1,000 and 2,500 ppm of carbon dioxide, with the order balanced across six groups and both the participants and the person administering the test blinded to the level. Ventilation rate and temperature were held constant, so the only thing that changed was the carbon dioxide, injected pure. At 1,000 ppm, six of the nine decision-making scales showed moderate and statistically significant decrements against 600. At 2,500 ppm, seven scales fell to between 0.06 and 0.56 of the 600 ppm score. One scale went the other way: focused activity rose, which is what someone does when they narrow down and stop taking in the wider situation. Satish asks for confirmation in the conclusion.
Xiaodong Cao exposed 15 people in a sealed chamber at 1,500, 3,500 and 5,000 ppm, each doing six cognitive tests covering perception, attention, short-term working memory, risky decisions and executive control. The result does not agree with Satish’s. Reaction time, speed perception and the 2-back memory test showed no significant difference at any level. Only at 5,000 ppm did response times lengthen, on visual search, the balloon risk task and the Stroop test. Cao’s recommendation is aimed at enclosed workplaces where a fast response matters, and it sets the threshold at a level four to five times higher than Satish’s.
Marc Syndicus ran two studies with 128 participants between them, one on sound and one on heat. In the first, 97 people made risky choices under three different sounds at 60 dB(A) and in quiet. Only one sound mattered, a radio podcast about a museum anniversary, and it made people more risk-averse rather than less. In a second study, 31 people did the same tasks either at 30 °C or above, or at 25 °C or below. In the warm room they made significantly riskier decisions on every task except the lottery choice, and the effect was stronger among the women in the sample.
Steffen Künn used a setting that scores its own decisions. Official chess tournaments monitor the air in the playing hall and every move can be scored objectively against an engine. Across 30,000 moves, each 10 micrograms per cubic metre more fine particulate matter indoors raised the probability of an erroneous move by 26.3 per cent. In that sample 10 micrograms is about three quarters of a standard deviation, so this is a difference between one ordinary tournament hall and another.
Piers MacNaughton takes the same decision-making test into an economic argument. Scores from an office with optimised daylight and views were ranked against a bank of more than 100,000 earlier results, which put workers at the 65th percentile against the 52nd for an office with traditional blinds, and the difference was converted through salary data into 352 billion dollars a year, or 1.7 per cent of US GDP. Four of the five authors also wrote the daylight experiment funded by View Inc., and one of them works for the company, which makes the electrochromic glass the argument recommends. The paper is behind a paywall here and the number of office workers behind the percentile shift is not in the abstract.
Three of the studies on this site use the same instrument. Satish’s nine decision-making scales, MacNaughton’s percentile and Joseph Allen’s cognitive scores on the temperature and air sheet all come from the Strategic Management Simulation, and Satish is an author on all three papers.
Satish worked with 22 people and Cao with 15, which is why they can disagree this sharply about the same gas. Syndicus ran 128 people across two experiments. Künn’s 30,000 moves are the only observations here from a real setting with something at stake, and MacNaughton’s figure is an extrapolation from a laboratory result to a national economy.