Winkler · Behaviour research and therapy 2025 · Prospective repeated-measures experimental analogue study · n=125

Associations among threat prediction error, prediction change, and anxiety during an exposure therapy analogue in adults with healthy to clinical social anxiety.

Level 3 - non-randomized controlled study

Prospective single-arm experimental analogue study with repeated within-subject measurements

PubMed 40023922 · doi:10.1016/j.brat.2025.104709 · record verified 2026-08-26

What was done

Undergraduate psychology students (N = 125) with social anxiety symptoms ranging from healthy to clinical completed an online exposure analogue consisting of 667 total speech performance exposures across two sessions separated by one week (~3 speeches per session). Self-reported anxiety, threat prediction, threat outcome, and surprise were recorded for each exposure. Multilevel modeling evaluated between- and within-person relationships between anxiety reduction and derived learning metrics (threat prediction error, prediction change, and prediction error learning rate).

What was found

No effect sizes, regression coefficients, or p-values were reported in the abstract. Between persons, average prediction change and prediction error learning rate were associated with greater anxiety reduction across exposures, whereas average prediction error itself was not. Within persons, anxiety was lower following exposures that elicited greater prediction change. Threat prediction error showed no linear association with anxiety at the subsequent exposure, but higher prediction error during an exposure was associated with higher subjective surprise.

Why it matters

This study refines inhibitory learning models of exposure therapy by indicating that generating large prediction errors may be insufficient for anxiety reduction unless those errors are actively converted into updated threat expectations.

Limits

The study used an undergraduate analogue sample and online speech tasks rather than clinical patients undergoing full exposure therapy courses. Specific numerical values and statistical test statistics were not provided in the abstract, and outcomes relied exclusively on self-reported ratings.

Cited by