It's meaningless to say cognitive scores "halve" or "double". That depends entirely on how the test results are presented, not the person or even the test. What

It's meaningless to say cognitive scores "halve" or "double". That depends entirely on how the test results are presented, not the person or even the test. What's half a human intellect? Is that being as smart as a dog? As smart as a drowsy person? Certainly not the difference between IQ 100 and IQ 50 because IQ is scaled so as to be normally distributed with 50 being almost impossible to achieve. Even the linked paper (great that it's linked and free access!) doesn't appear to explain what the scores mean in its "Cognitive Function Assessment" section but identifies the test as using "Strategic Management Simulation" software. So perhaps the conclusion should be "stuffy offices reduce cognitive function by an unknown amount".