Quick summary: Much of what psychology reports as a fact about people is also a fact about the room in which they were studied, from the lab and the survey culture to the ward and the workplace. A high burnout score, a psychiatric label that sticks, or a team that goes quiet in meetings can describe a reasonable response to staffing, rules, and power as much as a trait inside the person. For mental health practice and workplace policy, the useful move is to treat a finding as a hypothesis about a person in a setting, and to look at the room before assigning the problem to the individual.

Psychology likes to say it studies people. Look closer and a lot of it studies rooms: the lab, the ward, the survey platform, the culture that wrote the questions. Then it reports what it found as if the room had not been there.

Take the Müller-Lyer illusion, the two lines with arrowheads that look different lengths when they are not. It is a textbook staple, offered as proof of how human vision works. Yet when researchers tested people in different societies in the 1960s, susceptibility varied a great deal between groups. Joseph Henrich and colleagues pulled this and other cases together in their paper on WEIRD samples. Their point was blunt: the field has leaned heavily on participants from Western, educated, industrialised, rich, democratic societies, and then spoken as though they stood for everyone. The illusion was never about “the human eye”. It was about eyes raised among right angles.

Now bring that suspicion into the clinic.

In 1973, David Rosenhan published On Being Sane in Insane Places. Healthy volunteers reported hearing a voice, were admitted to psychiatric hospitals, then behaved normally. The label stuck anyway. Ordinary behaviour, such as note-taking, was read through the diagnosis. Later writers have challenged parts of the study’s methods, and that is worth knowing. But the uncomfortable question outlives the dispute: once a frame is in the room, how much of what we see in a person is the frame?

Michel Foucault pushed this further in Madness and Civilization, arguing that the modern asylum did not simply discover madness but organised it, sorting people according to what a society could and could not tolerate. You do not have to accept his whole case to take the lesson. A category is a tool. Tools are built by people for purposes, and they carry those purposes with them.

Here is the contrarian part. We are told that the cure for bias is better measurement. More validated scales, larger samples, cleaner statistics. These help. They also leave the room untouched. A burnout inventory handed to a nurse on a short-staffed ward measures her, and it also measures the ward, the rota and the hospital’s budget. When the score comes back high, we send the nurse to a resilience workshop. The ward goes on as before.

I see the same move in workplaces. A team goes quiet in meetings and we say the people are disengaged. Change the seating, who speaks first, or what happens to the last person who disagreed, and the “disengaged” people often talk. The trait was a response.

None of this means nothing is real. People suffer, and some suffering needs treatment. The claim is narrower and, I think, more useful: before you attach a trait to a person, ask what the room is doing. Who designed the test? Who was left out of the sample? What would this person look like somewhere else?

For students and practitioners, that points to three habits:

1. Ask who the finding was tested on, and who it was not.
2. Treat a diagnosis as a hypothesis about a person in a setting, not a fact about the person alone.
3. When a problem repeats across many people in one place, look at the place before the people.

A postmodern attitude toward psychology is often mistaken for cynicism. It can be the opposite. It treats knowledge as made by people, which means people can remake it. The most hopeful sentence in the field might be: this looks like a disorder, but what if it is a reasonable reply to a room that was built badly?

Jordan Bridger is a behavioural scientist with a PhD from the University Exeter and an Ivy League certificate in neuroscience. He founded HumanX, has run a consultancy for 15 years, and has written for HuffPost for over a decade.