Schedule of Events | Search Abstracts | Invited Symposia | Symposia | Poster Sessions | Data Blitz

Surprise? Surprise! Learned helplessness with/without action

Poster Session B - Sunday, March 8, 2026, 8:00 – 10:00 am PDT, Fairview/Kitsilano Ballroom

Jialin Shi¹ (jialinshi.ivy@gmail.com), Robin A. Murphy¹; ¹University of Oxford

Human perception of personal agency is based on statistical information (i.e. ΔP matrix, which records the frequencies of four types of events, A/B/C/D: binary states – presence vs. absence of action and outcome). Depression is characterised by reduced perceived agency, mirroring learned helplessness (LH). However, the classic LH paradigm often overlooks how the presence or absence of action modulates the learning of uncontrollability. Furthermore, while humans frequently infer control within social contexts, the mechanisms underlying controllability estimation in social settings remain poorly understood. To address this, we employed a contingency learning task where participants had no control over outcomes (ΔP = 0) and recorded event-related potentials (ERPs) time-locked to outcome delivery. In Experiment 1 (active learning), participants learned through direct experience. In Experiment 2 (observational learning), yoked participants learned by observing others. Analysis revealed: (1) During active learning, C events (outcome occurrence without action) elicited a P300 component with higher amplitude and increased latency compared to A events (concurrent action and outcome), suggesting prioritised processing for detecting control failures. (2) Active learning, compared to observational learning, was associated with increased P300 amplitude and prolonged latency overall. (3) Metacognitive assessments showed lower confidence in judgments of self-uncontrollability than other-uncontrollability. These results demonstrate that the learning of uncontrollability is fundamentally shaped by both action and social context. The neural and metacognitive dissociations between active and observational learning provide a novel framework for understanding how action and biased social learning may contribute to LH in depression.

Topic Area: EMOTION & SOCIAL: Emotion-cognition interactions