Explaining Too Much? Understanding How Large Language Model Reasoning Traces Influence Performance and Metacognition
2026-05-25 • Human-Computer Interaction
Human-Computer InteractionArtificial Intelligence
AI summaryⓘ
The authors studied how showing the steps a language model takes to arrive at answers affects people solving reasoning problems. They found that giving a simple summary of the model's reasoning made people trust the process more and enjoy it more, but didn't help them solve problems better. Showing the full detailed steps actually made performance worse. People also thought they did better than they really did, and none of the trace formats helped them accurately judge their own performance. The authors suggest these reasoning steps are more about how users experience the interface than revealing how the model actually thinks.
Large Language ModelReasoning TraceTransparencyLSAT-style ProblemsUser TrustPerformance CalibrationProcessing FluencyInterface DesignOverestimation BiasPreregistered Study
Authors
Daniela Fernandes, Daniel Buschek, Lev Tankelevitch, Thomas Kosch, Robin Welsch
Abstract
Large Language Model interfaces are increasingly verbose, exposing intermediate reasoning traces alongside final answers. Traces are framed as transparency mechanisms, yet it is unclear how people use them to solve problems. We report a preregistered between-subjects study (N = 559) in which participants solved ten LSAT-style reasoning problems under one of three conditions: an Answer-only baseline, a Full-trace revealed before the answer, and a Summary-trace presented alongside the answer. Summaries preserved task performance at the no-trace baseline while significantly elevating trust and hedonic appeal, establishing that trace exposure shifts subjective appraisal of the interaction without bringing performance benefits. Under an open-weight reasoning model exposing verbose intermediate output, full traces additionally impaired performance relative to the answer-only baseline. Across all conditions, participants substantially overestimated their performance, and no trace format supported calibrated self-evaluation. Further analysis indicates that hedonic appeal, not trust, carries the indirect path to overestimation, consistent with a processing-fluency account. Reasoning traces are best understood as user-facing interface artifacts rather than transparent windows into model cognition, and calibration is unlikely to emerge from the traces themselves and may best be scaffolded by interactions that elicit users' own reasoning first.