Thomas Gilovich walked a Cornell University undergraduate into a room full of other students, made him put on a bright yellow Barry Manilow T-shirt he had chosen specifically because Cornell students found it embarrassing, and then asked him to guess how many people in the room had noticed the shirt. The student guessed roughly half. The actual number, measured by asking each observer afterward, was closer to a quarter. Gilovich repeated the experiment across dozens of trials, and the pattern held so consistently that he and his co-authors gave it a name: the spotlight effect.
The finding was simple. People walking around wearing something conspicuous believe that about twice as many observers clocked it as actually did. The shirt was the vehicle, but the effect turned out to describe almost everything embarrassing a person can do in public.
The Barry Manilow shirt
Gilovich’s team ran the study at Cornell’s psychology department in Ithaca, New York. The setup was deliberately mundane. An undergraduate would arrive at the lab, be told to put on a T-shirt, and then be sent into an adjacent room where four to six other students were already seated filling out questionnaires. The wearer stayed for a few seconds — long enough to be seen, not long enough to strike up conversation — and then was pulled back out.
The shirt itself mattered. In pretesting, Cornell students had rated Barry Manilow as one of the most cringe-inducing celebrities they could imagine having their face plastered across their chest. That was the point. The researchers wanted an item the wearer would feel acutely self-conscious about, so that any gap between predicted and actual noticing would be as large as possible.
Wearers were then asked: of the people in that room, how many do you think could tell me who was on your shirt? Observers were asked the same thing in reverse: was there anything unusual about the person who just walked in, and if so, what?
The pattern was consistent: wearers predicted roughly twice as many observers had identified the face as actually did.
It wasn’t just embarrassing shirts
Gilovich anticipated the obvious objection — maybe Barry Manilow is a special case, maybe the effect only shows up when something is genuinely humiliating. So the team ran a second version with T-shirts that Cornell students had rated as things they would be proud to wear. Same result. Wearers overestimated how many observers had noticed the shirt by roughly the same factor.
The spotlight, in other words, wasn’t about shame. It was about the basic asymmetry between how much attention you pay to yourself and how much anyone else does. The wearer had been thinking about the shirt for the entire walk from the lab to the observation room. The observers had been filling out a questionnaire.
Further studies extended the effect to group discussions. Participants were asked to recall, afterward, contributions they had made — a comment that fell flat, a point they thought was clever. Then their fellow discussants were asked to rate how memorable each person’s contributions had been. Speakers consistently believed their remarks had landed harder, for better or worse, than listeners reported.
Why the numbers come out this way
The mechanism Gilovich proposed centered on how we estimate our salience to others. When you try to estimate how salient you are to other people, you start from your own vivid, first-person experience of yourself — the shirt is on your chest, the awkward comment came out of your mouth — and then adjust downward to account for the fact that others don’t share your perspective. The adjustment is real but insufficient. You never get all the way down to the correct number, because you can’t fully imagine what it’s like to be a bystander who has never met you and has their own errands to run.
Predicted noticing rates hovered around double the observed rates across conditions. The effect held whether the item was flattering or embarrassing, whether the observers were strangers or acquaintances, and whether the social act was wearing something or saying something.
The bad-hair-day cousin
Around the same time, Gilovich and Savitsky described a closely related phenomenon called the illusion of transparency — the belief that your internal states are leaking out through your face more than they actually are. If you feel disgusted by a drink you’re forced to sip politely, you assume observers can read the disgust on your expression. They mostly can’t. If you’re lying, you assume it shows. It usually doesn’t, at least not at the rate you fear.
Both effects come from the same root: the private, high-resolution feed you have of your own experience does not translate cleanly into what other people perceive of you. They are working with a much lower-bandwidth signal — a glance, a passing impression, a fragment of conversation — and they are, more often than not, thinking about themselves.
What the spotlight does to social anxiety
The spotlight effect is one of the more clinically useful findings to come out of late-1990s social psychology, because it maps directly onto the machinery of social anxiety. People with high social anxiety don’t just feel judged — they systematically overestimate the amount of judgment being directed at them. The stain on the shirt, the tremor in the voice, the flushed face: all of these register in the anxious person’s self-model as broadcast at full volume, when observers are catching perhaps a quarter of it, if that.
Clinicians who work with social anxiety and social phobia often use the Cornell finding directly. Telling a patient that Gilovich measured the ratio and it came out at roughly two-to-one is a more concrete intervention than telling them, generically, that they worry too much. The number gives the reassurance a spine.
The underlying anxiety machinery — the human capacity to imagine a future in which one is being watched and evaluated — is exactly what the spotlight effect exaggerates. You are running a simulation of how the room perceives you, and the simulation over-weights you because you are the only character in it whose interior you actually know.
Individual differences
Not everyone falls into the spotlight at the same intensity. People higher in trait social anxiety and self-consciousness appear to show larger gaps between predicted and actual noticing. Adolescents appear to show larger gaps than adults, which is consistent with the phenomenology of being fifteen and convinced the entire cafeteria has clocked your haircut.
The effect also plays out in dating and early-relationship contexts, where the stakes of being perceived feel highest. The spotlight can distort dating behaviour — people ruminate on a small verbal stumble from a first date that the other person has, in most cases, entirely forgotten by the following morning.
The Cornell ratio in a phone-camera world
The original experiment happened in a room with paper questionnaires and no phones. The ratio was clean because the observation window was seconds long and unaugmented. A quarter of a century later, the observation window has changed. If the Barry Manilow wearer walked into the same room in 2026, some fraction of observers might photograph the shirt, post it, and route it to strangers who would never otherwise have seen it.
Whether that changes the underlying psychology is a live question. Recent work on personality, social media exposure and anxiety in young adults suggests the spotlight has, if anything, been amplified for the cohort that grew up assuming any embarrassing moment could be captured and redistributed. The base rate of actually being noticed is still low in any given encounter — but the tail risk of a moment going viral, however remote, keeps the internal simulation running hotter.
The reassurance in the finding
What Gilovich and his co-authors gave, in the end, was a number. Not a metaphor, not an aphorism about how nobody is thinking about you — a measured ratio, with a control condition, in a peer-reviewed journal. Roughly half of the attention you think you are getting for the embarrassing thing is real. The other half is a projection you are running from inside your own head.
The finding has held up across replications and extensions, and it has been popularised widely enough that people who have never taken a psychology class use the phrase. The Barry Manilow shirt has been retired to the archive of famous stimuli, alongside the marshmallow, the Milgram shock generator, and the Stanford basement.
What lingers is the arithmetic. Observers notice far less than we predict. The next time the coffee spills down the front of a white shirt on the walk to the office, most of the people on the pavement will not have registered it, and among those who did, most will have moved on by the time they reach the next intersection. The spotlight is on, but the beam is narrower than it feels from inside the light.
