Users' baseline ranking accuracy significantly predicts their final ranking accuracy after AI collaboration, with no other significant effects¶

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 linear regression predicting final ranking accuracy, users' baseline accuracy was a significant predictor (b = 0.414, p < .001) while condition, time, and their interaction were not significant. → Cansu Koyuturk 2026

Evidence¶

Cansu Koyuturk 2026¶

Cansu Koyuturk, Sabrina Guidotti, and Dimitri Ognibene. (2026). The Hidden Cost of Contextual Sycophancy: an AI Literacy Intervention in Human–AI Collaboration. https://arxiv.org/abs/2605.18372 doi:10.48550/arxiv.2605.18372

q3 · i? · causal · r2

Linear regression on data from 60 participants in a mixed-design experiment predicting final ranking accuracy from baseline ranking, condition, time, their interaction, and scenario. The article reports "b = 0.414, SE = 0.075, z = 5.53, p < .001" for baseline accuracy, with all other effects non-significant (ps > .44).

"Users’ baseline accuracy significantly predicted final performance, b = 0.414, SE = 0.075, z = 5.53, p < .001, while no other effects were significant (all ps > .44)."

Discussion¶

Related Claims¶

• Users' baseline ranking accuracy significantly predicts the quality of the AI assistant's advice — related
• Carryover of user errors into AI advice reduces both advice quality and final ranking accuracy — related
• The sycophancy-focused intervention significantly reduced the assistant's positional mirroring of incorrect user rankings and rank-order alignment — related