Why confidence can drift away from being right
Confidence feels like a readout of how good the evidence is. But the brain is juggling two things at once: the quality of the sensory signal coming in, and the state of the body doing the judging. One account tested this by using pupil fluctuations as a window on unexpected arousal—brief shifts in bodily activation that aren't explained by the task itself. The results suggest arousal regulates how much perceptual precision counts, so that subjective confidence reflects a blend of available sensory signals and internal bodily states rather than accuracy alone [1].
That distinction matters because it means two people with identical evidence can feel differently confident, and the same person can feel more confident on one occasion than another for reasons unrelated to being more correct. In this framing, arousal isn't noise corrupting an otherwise clean accuracy signal; it's a second input that confidence quietly incorporates [1].
Feeling tuned in to your heartbeat is not the same as being tuned in
The heartbeat is the most-studied bodily signal in this area, and it exposes a gap between feeling and performance. In a study that had 66 participants track their heartbeat in short intervals twelve times a day across three days, researchers measured interoceptive metacognition—essentially, how well a person's confidence ratings matched how accurately they had actually counted their heartbeats. Only about 6% of the variation in this skill was attributable to stable differences between people, and that person-level component was measured reliably [2]. In other words, most of the mismatch between confidence and heartbeat-tracking accuracy shifts from moment to moment rather than being a fixed personal trait.
The same study found that people who rated themselves as more aware of their bodily experiences—on a standard self-report questionnaire, particularly its attention-regulation and self-regulation aspects—actually showed lower correspondence between confidence and accuracy [2]. The people most convinced of their interoceptive insight were, in this sample, the least calibrated. That is a caution about self-report: believing you notice your heartbeat is not evidence that your confidence about it is well grounded.
Where this shows up beyond the lab
The influence of bodily signals on confidence isn't confined to heartbeat counting. A study across the adult lifespan used a laboratory heartbeat counting task to measure interoceptive accuracy and metacognition—the latter captured through raw interoceptive confidence—and examined how these related to judgments of interpersonal trust [3]. This places bodily-signal monitoring in social decision-making, where confidence about one's read on another person could plausibly draw on internal state as well as outward evidence.
Clinical groups may show the same coupling with a different setting. In patients with somatic symptom disorder performing a heart rate discrimination task, participants reported lower decision confidence, and the way they process internal bodily signals is proposed as a possible contributor to disturbances in that processing [4]. Because this work compares patients within a group rather than isolating confidence from accuracy, it is best read as consistent with a body-to-confidence link rather than a demonstration that the link is independent of correctness [4].
What this means for trusting your own confidence
The practical upshot is that a surge of confidence is not a reliable certificate of accuracy, especially when it comes to reading your own body. Arousal can tilt confidence up or down while the underlying evidence stays the same [1], and self-rated bodily awareness can sit oddly with actual calibration—sometimes inversely [2]. If you want to know how well someone (including yourself) reads internal signals, compare their confidence against their measured performance rather than asking them how aware they feel [2].
What remains genuinely open is how far these effects generalize. The arousal finding comes from perceptual confidence [1]; the heartbeat work concerns interoceptive metacognition in daily life [2]; the lifespan study ties these measures to trust judgments [3]; and the patient study reports lower confidence in a clinical sample without cleanly separating it from accuracy [4]. Each supports a body-to-confidence influence in its own setting, but none establishes a single mechanism that operates identically across all of them.
Sources used in this answer
Unexpected arousal modulates the influence of sensory noise on confidence
Unexpected arousal, inferred from pupil fluctuations, appears to modulate confidence by combining sensory precision with internal bodily states, so confidence can shift independently of decision accuracy.
The misjudgment of interoceptive awareness: Systematic overrating of interoceptive awareness among individuals with lower interoceptive metacognitive skills
In a three-day ambulatory heartbeat-tracking study, only about 6% of variance in interoceptive metacognition was attributable to reliable between-person differences, and higher self-reported interoceptive awareness was associated with lower correspondence between confidence and accuracy.
Interoceptive Accuracy, Metacognition, and Interpersonal Trust Perception Across the Adult Lifespan
Interoceptive accuracy and confidence-based metacognition, measured with a laboratory heartbeat counting task, were examined in relation to interpersonal trust judgments across the adult lifespan.
Within-group variation in interoceptive bias between patients with somatic symptom disorder on a heart rate discrimination task
Patients with somatic symptom disorder on a heart rate discrimination task reported lower decision confidence, with altered processing of internal bodily signals proposed as a possible contributor.