Research gap at a glance
- Question: In one unselected sample, does interoceptive accuracy predict felt arousal as a main effect, only as a moderator of the bodily response, or both?
- Why it matters: Jamesian accounts predict that better perceivers feel more; hybrid two-factor accounts predict that accuracy gates arousal only through the bodily response. The two existing studies disagree in a way that sampling design alone could explain, so the field’s picture of how interoception shapes emotion rests on an unresolved conflict.
- Theme: Emotion and self
- Status: open
What the wiki already knows
- In 44 adults selected as good or poor heartbeat perceivers, good perceivers rated affective pictures as more arousing (r = 0.34 with the heartbeat score) and showed larger P300 and slow-wave amplitudes (Pollatos et al. (2005)).
- In an unselected sample of 58, interoceptive accuracy did not predict arousal ratings, but it moderated the link between heart-rate response and felt arousal (Delta R squared = .12) and not valence (Dunn et al. (2010)).
- The same pattern of selected samples finding main effects and unselected samples not finding them recurs in decision-making, where sampling inflation and mixture purification both predict it (Does somatic feedback guide decisions?, Werner et al. (2009)).
- Among accurate heartbeat detectors, good detectors felt emotion more intensely with sympathetic arousal controlled (Wiens et al. (2000)).
- Single-session accuracy scores carry substantial within-person variance, which attenuates correlations of the size at issue (about r = .08 to .34) by an amount nobody in the dispute has computed (State vs trait interoception).
What would answer it
- Measure interoceptive accuracy (counting and discrimination), heart-rate response during Affective picture viewing and trial-wise arousal and valence ratings in one unselected sample of several hundred, estimate the main effect and the accuracy by bodily-response interaction together, and then draw extreme-groups subsamples from the same data to test whether selection alone reproduces the Pollatos main effect.
Connections
- Raised on: Is more interoceptive awareness better?, Pollatos et al. (2005), Dunn et al. (2010), State vs trait interoception
- Related concepts: Core affect, Interoceptive sensitivity, Cognitive appraisal (of emotion)
- Related debates: Is more interoceptive awareness better?, Is the heartbeat-counting task a valid measure of cardiac perception?, Does somatic feedback guide decisions?
Is more interoceptive awareness better? names the adjudication of the Pollatos and Dunn disagreement as its most tractable open question, because the two designs differ in only a few identifiable ways. Pollatos et al. (2005) selected extreme groups and had no bodily-response term. Dunn et al. (2010) recruited an unselected sample, measured the bodily response, and found no main effect but an interaction.
Each design leaves a hole the other fills. Pollatos’s sample may inflate a main effect through selection, or purify it if accurate perceivers are a distinct minority (see Is heartbeat-counting performance a continuum or a mixture of two groups?). Dunn’s sample may have been too small to detect a main effect of the size Pollatos’s design suggests. The decisive study puts accuracy, the bodily response and felt arousal in one unselected sample large enough to estimate both terms.
The answer also bears on theory. Dunn et al. found the moderation for arousal but not valence, which they read as support for hybrid two-factor accounts over strong Jamesian ones (Cognitive appraisal (of emotion), Core affect). If a main effect turns out to be real in a large unselected sample, the Jamesian reading regains ground.