Hedonic appeal, not trust, carries the indirect path from trace exposure to performance overestimation, consistent with a processing-fluency account¶

Claim · All claims
Evidence · 1 study · causal r2 · q3
Currency · newest study 2026 · not yet checked against its sources · not reviewed by a person

Subclaims¶

q3 i? In a parallel multiple-mediator model, both trace conditions raised trust and hedonic appeal, but only hedonic appeal predicted overestimation once trust was partialled out, so the indirect path through hedonic was significant for both conditions while the trust path was not. → Fernandes 2026

Evidence¶

Fernandes 2026¶

Fernandes, D., Buschek, D., Tankelevitch, L., Kosch, T., & Welsch, R. (2026). Explaining Too Much? Understanding How Large Language Model Reasoning Traces Influence Performance and Metacognition. https://arxiv.org/abs/2605.25856 doi:10.48550/arxiv.2605.25856

q3 · i? · causal · r2

Parallel multiple-mediator model in lavaan (N = 559, 5,000 percentile bootstrap samples) with condition as predictor, trust (TSQ) and UEQ hedonic as parallel mediators, and overestimation bias as outcome; the hedonic b-path was b2 = 0.15, p = .007 while the trust b-path was not significant (b1 = 0.09, p = .13).

"Both trace conditions raised trust (𝑎1,full = 0.26,𝑝=. 011;𝑎1,summary = 0.30,𝑝=. 005) and hedonic appeal (𝑎2,full = 0.45;𝑎2,summary = 0.45; both 𝑝<. 001), but only hedonic appeal predicted overestimation once trust was partialled out (𝑏2 = 0.15,𝑝=. 007;𝑏1 = 0.09,𝑝=. 13)"

Discussion¶

Related Claims¶

• Reasoning-trace exposure raised trust and hedonic appeal without performance benefits, shifting subjective appraisal of the interaction — related
• Perceived trust mediates the relationship between social influence and generative AI dependence — related
• Participants substantially overestimated their performance in all trace conditions, with the largest overestimation under full traces — related
• Summary traces preserved task performance at the answer-only baseline while elevating subjective appraisal — related
• Full reasoning traces impaired task performance relative to both answer-only and summary-trace conditions in AI-assisted LSAT reasoning — related