The Neuroscience of Confidence: How Performance Becomes Belief.

Most people treat confidence as a character trait, an endowment distributed unevenly at birth, or a quality that develops slowly through decades of accumulated life experience. That framing is wrong, and it matters that it is wrong, because it leads to exactly the wrong interventions. If confidence is a trait, the prescription is patience, positive thinking, or perhaps a motivational seminar. If confidence is something else, a neurological computation, a prediction generated from specific inputs, then the prescription changes entirely.

In 1977, Albert Bandura proposed that something else. He introduced the concept of self-efficacy: your brain's prediction of whether you can successfully execute a specific action in a specific context.^[1] This was not confidence in the colloquial sense. It was not self-esteem, which is a global evaluation of self-worth. It was not optimism, which is a disposition toward positive outcomes. Self-efficacy is granular, task-specific, and updatable. Your brain generates it from evidence, and the evidence is performance.^[2]

That distinction turned out to be one of the most productive ideas in the history of behavioural science. Over the next five decades, Bandura's framework was tested across more than 2,500 empirical studies,^[2] replicated in domains from surgery to sales, and eventually traced to a specific neural pathway running between the brain's reward-processing system and its self-attribution centres.^[10]

The practical consequence of this reframing is uncomfortable. If confidence is a prediction based on evidence, then people who lack it are not weak or defective. They are undersampled. They have not fed their brain enough of the right kind of performance data, or their brain's updating mechanism has misfired, discounting successes and amplifying failures.

This is precisely what happens in impostor syndrome. Pauline Clance and Suzanne Imes first described the impostor phenomenon in 1978: high-achieving women who attributed their success to luck, timing, or the goodwill of others, despite overwhelming objective evidence of competence.^[7] A systematic review of 62 studies covering 14,161 participants found prevalence ranging from 9% to 82% depending on the population and measurement instrument.^[8] Among healthcare professionals specifically, a 2025 meta-analysis of 30 studies calculated a pooled prevalence of 62%.^[9] These are not people who lack ability. They are people whose self-efficacy updating mechanism has broken: the brain discounts positive evidence and overweights negative signals.

The gap between competence and confidence has measurable economic consequences. Among STEM graduates, Sterling and colleagues found that the confidence gap, not a competence gap, not a values gap, mediated a $1,996 annual starting salary difference between men and women.^[18]

Bandura identified four sources of self-efficacy, and they are not equally powerful.^[1] Enactive mastery experience (direct personal performance) is the strongest. Vicarious experience (watching someone similar succeed) is second. Verbal persuasion (being told you can do it) ranks third. And physiological arousal (interpreting your body's state as evidence of capability or incapability) is fourth.^[2] In physical activity research, consistent with Bandura's own theoretical prioritisation, mastery experience consistently ranks as the most potent source of self-efficacy.^[40]

The hierarchy matters because it tells you what the brain is actually listening to. Someone can receive verbal encouragement every day, "You're brilliant, you can do this," and still feel like an impostor, because the brain weights direct performance evidence more heavily than social signals. Verbal persuasion may produce short-term increases in self-efficacy, but those gains remain vulnerable to erosion from subsequent failure experiences.^[2] The brain wants to see you do the thing, not hear that you can.

A 2022 fMRI study by Shany and colleagues at Tel Aviv University revealed why. When participants received positive social feedback about their performance, the ventral striatum, the brain's core reward-encoding region, activated in proportion to their subsequent self-efficacy updating.^[10] The ventral striatum generated what neuroscientists call a prediction error signal: the difference between what the brain expected and what actually happened. Positive prediction errors, outcomes better than expected, drove self-efficacy upward.

The circuit does not stop at the ventral striatum. Shany's team identified a specific connectivity pathway between the ventral striatum and the posterior middle temporal gyrus (pMTG), a region involved in social cognition and self-referential processing.^[10] The strength of this VS–pMTG connectivity predicted the degree to which participants showed a positive bias in self-efficacy updating. Participants with stronger connectivity updated their beliefs more optimistically after success. Those with weaker connectivity, who also scored higher on measures of anxiety, depression, and low self-esteem, showed blunted updating.

This is the neural architecture of impostor syndrome. The reward signal fires, the person succeeds, but the attribution circuit fails to route that success to the self-model. The prediction error is generated, but it never updates the belief. The medial prefrontal cortex (mPFC), which encodes self-relevant evaluations, responds differently depending on existing self-esteem. Somerville and colleagues demonstrated that individuals with low self-esteem showed exaggerated mPFC responses to evaluative social feedback: they were not ignoring feedback but over-processing it, in a destabilising way.^[11]

That matters because impostor syndrome is not an absence of signal. It is a corruption of signal routing. The evidence arrives, but the brain sends it to the wrong address.