Reasoning-trace exposure raised trust and hedonic appeal without performance benefits, shifting subjective appraisal of the interaction¶

Claim · All claims
Evidence · 1 study (2 entries) · causal r2 · q3 · i1–i2
Currency · newest study 2026 · not yet checked against its sources · not reviewed by a person

Subclaims¶

q3 i1 Trust in the AI differed significantly across conditions, with Answer-only participants reporting lower trust than both trace conditions. → Fernandes 2026 q3 i2 Hedonic quality on the UEQ-S was rated significantly higher in both trace conditions than in Answer-only. → Fernandes 2026 (2)

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 · i1 · causal · r2

XAI Trust scale ANOVA was significant with η² = .018; Tukey-adjusted comparisons showed Answer-only (M = 3.61) lower than Full-trace (M = 3.80, padj = .030) and Summary-trace (M = 3.83, padj = .011), with no Full- vs Summary difference.

"Trust in the AI differed significantly across conditions(𝐹( 2, 556)= 5.00,𝑝=. 007,𝜂 2 =. 018)."

Fernandes 2026 (2)¶

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 · i2 · causal · r2

UEQ-S hedonic quality ANOVA was significant (F(2,556)=13.31, p<.001, η² = .046); both trace conditions exceeded Answer-only while not differing from each other, indicating verbosity beyond summaries added no subjective-experience benefit.

"For Hedonic Quality, bothFull-traceandSummary-tracewere rated significantly higher thanAnswer-only(both 𝑝𝑎𝑑𝑗 <. 001), with no significant difference betweenFull-traceandSummary-trace( 𝑝𝑎𝑑𝑗 =. 999)."

Discussion¶

Related Claims¶

• Full reasoning traces impaired task performance relative to both answer-only and summary-trace conditions in AI-assisted LSAT reasoning — related
• Hedonic appeal, not trust, carries the indirect path from trace exposure to performance overestimation, consistent with a processing-fluency account — 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 worsened calibration bias and Brier scores relative to both other conditions — related
• No trace format supported calibrated self-evaluation: perceived-achieved correlations were low and non-significant in all conditions — related