Interoceptive accuracy
How well a person can actually detect the signals coming from inside their own body, measured by how their performance on a task lines up with what is really happening — for instance, judging whether a tone falls in time with their own heartbeat. It is the objective score of bodily self-detection, and it does not have to match how attuned a person feels.
The objective behavioural-task measure of how well a person can detect their own interoceptive signals — most commonly operationalised through the heartbeat-detection task, in which the subject judges whether external tones (or button-presses) are synchronous with their own heartbeats. Accuracy is the percentage match between perceived and actual cardiac events, treated as a stable individual-difference variable that ranges across normal populations from chance-level (~50%) to near-perfect detection.
Etymology§
The construct emerges from the psychophysiology literature of the 1970s and 1980s — Jones, Schandry, Whitehead and colleagues working on heartbeat perception as an experimental paradigm — but the term interoceptive accuracy in its current narrow sense is from Garfinkel et al. 2015, where it is one of three orthogonal constructs disambiguated from earlier loose usage. Pre-2015 papers calling this dimension interoceptive awareness are using awareness in the older conflated sense.
Interoceptive accuracy is the dimension of interoception that has been most measurable, most often measured, and (until 2015) most often confused with other dimensions. The heartbeat-detection paradigm — judge whether the tone is synchronous with your heartbeat — produces a single number per subject that has been correlated, across hundreds of studies since Critchley et al. 2004, with right anterior insula activity and grey-matter volume, with emotional-experience intensity, with anxiety and depression measures, and with eating-disorder and addiction-vulnerability outcomes.
Accuracy is not awareness. A person can be objectively accurate without subjectively knowing they are accurate; another can feel highly attuned and be objectively poor. The Garfinkel 2015 paper showed the two dimensions dissociate in a normative sample of 80 subjects, with sensibility (subjective self-report) as a third orthogonal axis. The post-2015 clinical literature uses the three-way distinction routinely; pre-2015 results have to be re-read with the distinction in mind to know which construct each study actually measured.
The heartbeat-detection task itself is methodologically contested. Brener and Ring 2016 in Philosophical Transactions of the Royal Society B documented that subjects' beliefs about their own heart rate predict counts more reliably than the actual heartbeats do; medical professionals and athletes, who know their own typical heart rate, score better not because of better interoception but because of better priors. The heart-rate discrimination task developed by Legrand, Allen and colleagues at Aarhus from 2021 on is one response; the cardiac elevation-detection task is another. Interoceptive accuracy is a real but partially confounded construct, with a methodology actively correcting itself in mid-flight.