We remember first impressions as photos, but we actually build them from video
Ask someone to picture a first impression and most people imagine a single face — a still, front-facing face, like an ID photo. But in real life, the moment you meet someone is never a freeze-frame. The stride through the door, the way they take a seat, where the hands come to rest, the pitch and speed of the first words — a first impression is made as all of this movement tangles together within a few seconds. If the earlier articles about still face photos explained the 'raw materials' of impression, this one looks at what happens when those materials start to move.
Let me say this up front. The psychology studies introduced here do not claim that you can 'read off' a person's character, ability, or destiny from a face or a gesture. If anything, they show the opposite: how quickly and how confidently we manufacture impressions, and how far those impressions can drift from who the person actually is. This is not a technique for judging people but a culture-and-mind story that tries to understand the moment a judgment is formed. This article does not judge anyone's real personality, ability, health, or destiny from a face or a body.
People reach conclusions from just a few seconds — a 'thin slice'
In 1992, psychologists Nalini Ambady and Robert Rosenthal published a meta-analysis that crystallized the concept of 'thin slices.' Synthesizing many studies, its central finding was that even when a person's expressive behavior is cut down to a very short clip — sometimes under 30 seconds — the ratings observers give are strikingly consistent with one another and line up reasonably well with certain external outcomes. For example, the ratings of people who watched only a few silent seconds of a teacher's class often pointed in a similar direction as the evaluations of students who had spent a whole semester with that teacher.
More striking was the matter of time. Extending the observation to several minutes did not make judgments much more accurate than a brief observation of under five minutes. In other words, rather than gathering information over a long stretch and concluding carefully, we largely hold onto the impression formed in the first few seconds. But do not misread the word 'accurate' here. What Ambady and Rosenthal showed was 'consensus' — that observers see things similarly — and 'association' — that the impression predicts certain outcomes to a degree. It does not mean the short slice sees through to the other person's true inner self.
Judging a stranger — some traits read clearly, the rest stay blurry
High consensus does not mean high accuracy. The German psychologists Peter Borkenau and Anette Liebler, also in 1992, studied personality inference at 'zero acquaintance.' Observers who had never once spoken with the target were given only a short video or a voice recording and asked to guess a stranger's personality. The results ran in two directions. Observers shared quite similar impressions (consensus was high), but the degree to which those impressions actually matched the target's self-report or the ratings of acquaintances — that is, accuracy — varied greatly from trait to trait.
In particular, traits tied to outwardly visible behavior — how talkative someone is, how loud the voice, how lively the movement — such as extraversion, even strangers guessed relatively well. By contrast, traits that do not surface easily in appearance, like conscientiousness or emotional stability, saw observers confidently converge on an opinion while that opinion diverged from the person themselves. In short, 'everyone sees it that way' and 'it is actually so' are entirely different matters. The confidence of a first impression is not evidence of its accuracy.
The voice is read as fast as the face
A thin slice need not be something seen with the eyes. Among the studies Ambady and Rosenthal synthesized were some that stripped away expression and gesture and left only the pitch and rhythm of the voice, and even from such auditory information alone observers formed fairly consistent impressions. The speed of speech, where the pauses fall, whether sentences rise or fall at the end — these prosodic textures shape a feeling about the person quite apart from the content of the words. In Borkenau and Liebler's experiments too, the voice was one of the important channels for judging a stranger.
What is interesting is that the voice often arrives before the face. A phone call, laughter leaking out of a meeting room, a conversation drifting from behind you in a café — we have already drafted an impression from the voice before we ever see the face. But the same caution applies here. A low, slow voice does not amount to a calm personality, nor a fast, high voice to an unstable one. The voice is merely a surface that shifts freely with the day's condition, the situation, and cultural speech habits — it cannot be grounds for pinning down the person underneath.
A body standing still and a body standing as itself tell different stories
A 2009 study by Laura Naumann and colleagues extended this difference to the whole body. The researchers showed people full-body photographs and had them infer personality, but split the setup into two conditions. One was a 'controlled' pose in which expression was neutralized and everyone stood the same way with arms hanging down; the other was a 'natural' pose in which each person stood however they liked. In the controlled pose, only extraversion and a sense of self-esteem could be read to any degree.
But once people stood in a posture of their own choosing, the range of traits observers could guess to some degree widened noticeably. Whether the arms are crossed, whether the body opens toward the camera, where the center of gravity rests — such spontaneous postures held more cues actually connected to the person. The implication is clear. What sways a first impression is less the inherited arrangement of features than a 'habit of movement' — how the person uses their body. And that habit is a matter on a different layer from the face itself.
When the face cannot speak, the body speaks in its place
The study that showed the relationship between face and body most dramatically was a 2012 experiment by Hillel Aviezer, Yaacov Trope, and Alexander Todorov. The researchers gathered photos of professional tennis players at the moment they won a decisive point and the moment they lost one. When only the faces were cropped and shown, people could barely tell whether it was an expression of elation or of despair. At the peak of intense emotion, the face did not send as clear a signal as we assume.
The twist came when the body was shown together. The same face, placed on a winning body, was read as joy; placed on a losing body, as agony. When the researchers deliberately composited a 'winning face' onto a 'losing body,' people's judgments followed the body, not the face — even while believing they had judged from the face. The more intense the moment, the more we read emotion from posture and gesture rather than from the face. The common notion that 'it's all written on the face' was, at least in such moments, not true.
| Type of cue | Observer consensus | Accuracy vs. actual personality (not a basis for judgment) | Related study |
|---|---|---|---|
| Still face / features | High | Limited — blurry especially at intense emotion | Aviezer et al. (2012) |
| Voice prosody (pitch, speed) | High | Relatively for extraversion-type traits, low for the rest | Ambady & Rosenthal (1992) |
| Controlled (forced) posture | Moderate | Only extraversion and self-esteem to a degree | Naumann et al. (2009) |
| Natural (spontaneous) posture / gesture | High | Widens across several traits but still partial | Naumann et al. (2009) |
| Overall impression at zero acquaintance | High | Large variance by trait — consensus does not equal accuracy | Borkenau & Liebler (1992) |
The distance between the perceived impression and the actual person
One distinction runs through all these studies: the distance between the 'perceived cue' — what people feel to be the case — and the 'actual personality' — who the person really is. Thin slices, zero-acquaintance judgments, posture, gesture — all of these leave a strong impression on the observer. But a strong impression is not itself true information. Because observers agree with one another, it is easy to mistake that consensus for 'objective fact,' yet the studies above show, again and again, that the consensus often diverges from the person's actual self.
On top of this, first impressions carry the risk of feedback. If you assume someone is a lively person, you find yourself talking to them more, and then they really do respond with liveliness. Your prediction draws out the other's behavior and confirms itself. So a first impression is less a matter of 'reading something out' than an interaction that observer and target create together. No cue can pin down that person's character or ability, let alone their destiny.
So how should we treat first impressions?
The practical conclusion of all this is unexpectedly modest: a first impression is not a conclusion but a starting point of information. The stronger the certainty formed in a few seconds, the better it is to leave open the possibility that it is only a consensus impression, not a verified fact. Especially in settings where judgment affects someone's life — hiring, introductions, meetings with strangers — it helps to treat the first few seconds' impression as a 'hypothesis' and update it with the person's actual later behavior.
Seen from the side of managing one's own impression, there is comfort in how much room there is to change — more than you might think. This is not about altering inherited features. The cues the studies point to lie mostly in the realm of movement and bearing: the speed of the voice, the direction of the gaze, the tension in the shoulders, the angle at which the body opens. This is not a technique for deceiving others but a matter of communication — conveying what you want to convey a little more accurately. And nothing in this article can serve as grounds for measuring a person's worth by face or body. What it holds is not a tool for judgment but the humility of knowing how easily judgment goes wrong. Enjoy it as a culture-and-mind story to read for fun, and please do not use it where people are being pinned down.
Frequently asked questions
Are first impressions really settled in just a few seconds?
Several studies find that people form consistent impressions from very brief observation. In Ambady and Rosenthal's 1992 meta-analysis, extending the observation time did not make ratings much more accurate than a brief observation of under five minutes. But 'formed quickly' and 'accurate' are different matters, and it is safer to treat a first impression as a hypothesis rather than a fact.
Can a face photo alone reveal someone's personality? (It does not.)
You cannot pin it down. According to Naumann and colleagues' 2009 study, only extraversion could be read to a limited degree in a controlled pose with expression neutralized, and it was only in a natural posture that more cues emerged. In other words, movement and posture sway the impression more than static features, and even then it is hard to say you 'read off' personality. This is a culture-and-mind perspective, not a factual verdict about a person.
Is it true that 'emotion is all written on the face'?
Not always. In Aviezer, Trope, and Todorov's 2012 experiment, when only the face at the peak of intense emotion was shown in isolation, people could not even distinguish elation from despair, and it was the posture of the body that led their judgment. It means that the more intense the moment, the more information the body conveys rather than the face.
If everyone feels a similar first impression, doesn't that make it right?
Consensus and accuracy are different. In Borkenau and Liebler's 1992 study, observers shared similar impressions, but how well those impressions matched the person's actual personality varied greatly by trait. The fact that many people received the same impression does not make that impression true.
So to make a good first impression, do I have to change my face?
No. The cues the studies point to lie mostly in the realm of movement and bearing — voice, gaze, posture, the angle at which the body opens. This article is not medical advice and does not recommend any procedure or surgery. An impression is less a matter of fixing the face than of communication — conveying what you want to convey more accurately.
Article info & references
Published July 4, 2026 · Last updated July 4, 2026
- Ambady, N. & Rosenthal, R., Thin Slices of Expressive Behavior as Predictors of Interpersonal Consequences: A Meta-Analysis (1992)
- Borkenau, P. & Liebler, A., Trait Inferences: Sources of Validity at Zero Acquaintance (1992)
- Naumann, L. P., Vazire, S., Rentfrow, P. J. & Gosling, S. D., Personality Judgments Based on Physical Appearance (2009)
- Aviezer, H., Trope, Y. & Todorov, A., Body Cues, Not Facial Expressions, Discriminate Between Intense Positive and Negative Emotions (2012)