In an experiment engineered to draw maximum attention, more than three quarters of observers didn't notice. The psychology of why you feel watched - and the data showing you aren't.
In the late 1990s, a psychologist at Cornell handed a student a T-shirt with a large photograph of Barry Manilow on the front, told him to put it on, and sent him into a room where four or five other students were filling out questionnaires.
Moments later the experimenter pulled him back into the hallway and asked one question: how many of the people in that room could tell you who was on your shirt?
The answer he gave, and the answer the actual observers gave, is one of the most useful findings in social psychology — and it applies with uncomfortable precision to every changing room you've ever hurried through.
Thomas Gilovich, Victoria Husted Medvec and Kenneth Savitsky published the study in the Journal of Personality and Social Psychology in February 2000. They recruited 109 Cornell undergraduates; fifteen of them wore the Manilow shirt, chosen specifically because the researchers had established it was maximally embarrassing for that population.
The wearers estimated, on average, that 46% of the people in the room could identify the face on their chest.
The actual figure was 23%.
The researchers named it the spotlight effect: our tendency to believe the spotlight shines on us far more brightly than it actually does.
The obvious objection is that embarrassment specifically distorts things. So they ran it again with shirts people were happy to wear — Bob Marley, Jerry Seinfeld, Martin Luther King Jr.
Wearers estimated 48% would identify the face. The actual figure was 8%.
A wider gap than the embarrassing version. The bias isn't about shame — it's structural. It applies to anything you're aware of about yourself, good or bad.
Two studies, same direction of error. People consistently forecast an audience that isn't there.
Gilovich's explanation is called anchoring and adjustment, and once you see it you can't unsee it.
You experience yourself from the inside, in extremely high resolution. You know exactly what you're wearing, exactly how you're standing, exactly what you're worried about. That internal experience is vivid, continuous, and impossible to switch off.
When you try to estimate how you appear to others, you start from that vivid internal picture and adjust downward. And the adjustment is real — you do know intellectually that others see less. But you don't adjust nearly far enough, because you can't fully imagine an experience where you're not the main subject.
Meanwhile every other person in the room is doing exactly the same thing about themselves.
Gilovich ran a control group who watched footage of the same scenarios and estimated how many observers would notice. Their estimates were far lower — close to the real figure.
The inflation wasn't a general belief that people are observant. It happened specifically to the person who was the subject. Being the one it's happening to is what produces the error.
The same research group documented a second, related bias — the tendency to believe your internal states are visible from outside.
In one study, participants tried to conceal disgust while tasting an unpleasant drink. They were confident observers could tell. Observers mostly could not.
Apply that to a changing room. The feeling of being self-conscious is intense, and it feels like it must be radiating off you — like everyone can see you trying not to be seen. The evidence says they can't. Your nervousness is far less visible than it feels, which means the loop of being anxious about looking anxious is running on a false premise at every level.
Everyone looked. Everyone formed an opinion. They'll remember. They can tell I'm uncomfortable, which makes it worse.
Roughly half as many people noticed as you'd predict — and in the flattering-shirt study, six times fewer. Internal states are substantially less visible than they feel. Observers were mostly occupied with themselves.
The Manilow shirt was designed to be maximally attention-grabbing: a large, high-contrast image on a person's chest, in a room of peers, deliberately selected for social awkwardness. Under those engineered conditions, 77% of people still failed to register it.
Now consider the situations you're actually worried about. A changing room where everyone is facing their own locker and moving fast. A urinal where looking is itself the taboo. A pool. A shared shower. In every one of those, the social norm is explicitly not to look — which means the baseline attention is far lower than in Gilovich's deliberately provocative setup.
You're worried about being observed in the environments humans are most careful about not observing each other.
Being straight about the limits: these were controlled experiments, mostly with university students, measuring whether observers could recall a shirt or a comment. They don't prove that nobody ever notices anything or that people never judge. The consistent, replicated finding is narrower and still useful — the direction of the error. People reliably overestimate their own noticeability. The size of the overestimate varies by situation. The direction doesn't.
When you catch yourself calculating how many people noticed, cut the number in half and then cut it again. You will still probably be high. This isn't positive thinking — it's correcting for a measured, replicated bias in the same way you'd correct a scale you knew read heavy.
Think about the last changing room you were in. Can you describe anyone else's body in it? Almost certainly not — and you were physically present, with the same eyes and the same opportunity. That is exactly how much they retained about you.
Rushing, angling away, avoiding, leaving early — behaviour is far more noticeable than anatomy. Ironically, the thing most likely to draw attention is the effort to avoid it. Moving normally is genuinely the most invisible option available to you.
You are the main character of exactly one life. In everyone else's, you are a background figure they will not recall by dinner.
That sounds bleak written down. In practice it's one of the most freeing facts available — because it means the audience you've been performing for, defending against, and shrinking in front of has been almost entirely constructed inside your own head.
The other thing worth correcting: what you think "average" actually means. Real clinical data, not locker-room estimates.
See the Real Numbers →Fifteen students walked into a room with a giant photograph of Barry Manilow across their chests, in an experiment built to maximise attention, and more than three quarters of the people present didn't register it.
Whatever you're currently worried someone noticed about you — under clothes, in a room where looking is taboo, for a few seconds — is a far smaller signal than that shirt.
Nobody was looking. They were busy assuming you were looking at them.