When you feel like “everyone saw me make a fool of myself,” cut that estimate in half

When walking into a room wearing an embarrassing T-shirt, people assume roughly half of onlookers recognize who is on the shirt, when in reality fewer than a quarter notice. They similarly overestimate when wearing a respectable T-shirt. After group discussions, people overestimate how noticeable their verbal missteps were to others, and after casual chats, people consistently underestimate how much the other person liked them. When you feel like “everyone saw me make a fool of myself,” the actual number of people who noticed is typically only half or less.

No cost at all. It simply requires asking yourself once, when you’re nervous: “Did they really see it?”

Experiment 1: Participants donned T-shirts featuring Barry Manilow’s portrait and entered rooms with other subjects. On average, they predicted that 46% of observers would correctly name the figure on the shirt; the actual correct rate was 23%, precisely half the estimate. Subtracting the observed rate from the predicted one yields an average over-estimation of 23 percentage points (95% CI 9–38). Excluding two sessions with only two or three observers leaves a difference of 27 points (11–43). Experiment 2: Participants wore T-shirts they personally selected as "positive." Their average prediction was 48%, whereas the true recognition rate fell to 8%. Across three additional contexts—strangers paired for conversation, college roommates, and adult workshop participants—people systematically underestimated both how much their interlocutor liked them and how much they enjoyed the talk. The authors term this the "liking gap." It appears in conversations of any length and persists for months among new roommates.

Grade B reflects that all studies were conducted in laboratory or campus settings with relatively small sample sizes, involving wearing shirts or chatting with strangers. There is no direct evidence that these findings apply to situations such as being discussed by colleagues or making a mistake in a meeting. The numbers merely indicate the direction and approximate magnitude of overestimation; they do not provide a fixed multiplier you can apply. This advice is not meant to imply that others' opinions are irrelevant. In genuine evaluation contexts—interviews, defenses, assessments, or medical evaluations—judgments are real and should be taken seriously. It specifically addresses that everyday feeling of "everyone is watching me"; the real opponent here is your own estimation bias.

Gilovich T, Medvec VH, Savitsky K (2000). The spotlight effect in social judgment: An egocentric bias in estimates of the salience of one's own actions and appearance. Journal of Personality and Social Psychology, 78(2), 211–222. https://doi.org/10.1037/0022-3514.78.2.211；Boothby EJ, Cooney G, Sandstrom GM, Clark MS (2018). The Liking Gap in Conversations: Do People Like Us More Than We Think? Psychological Science, 29(11), 1742–1756. https://doi.org/10.1177/0956797618783714