Proactive AI agent dialogues showed stronger conceptual–adequate reasoning links while reactive dialogues showed stronger factual/off-task associations¶

Claim · All claims
Evidence · 1 study · causal r2 · q2
Currency · newest study 2025 · not yet checked against its sources · not reviewed by a person

Subclaims¶

q2 i? Conditions separated along MR1 (W=228, p<.001, Cliff's δ=−0.56, large), indicating different relational organisation of reasoning. → Yueqiao Jin 2025

Evidence¶

Yueqiao Jin 2025¶

Yueqiao Jin, Kaixun Yang, Roberto Martinez-Maldonado, Dragan Gašević and Lixiang Yan. (2025). The Agency Gap in AI-Supported Writing: How Reactive and Proactive Agent Designs Shape Multimodal Reasoning. https://arxiv.org/abs/2507.04398 doi:10.48550/arxiv.2507.04398

q2 · i? · causal · r2

Mann–Whitney comparison of ENA condition centroids in the same discourse-network model found large separation (Cliff's δ=−0.56); the reactive side's stronger factual/off-task association (EP-PS–EP-OFF, w=0.27) is reported in the same section.

"Reactive and proactive human–AI interaction designs were separated along MR1 (𝑊=228,𝑝<.001, Cliff’s𝛿=−0.56, large), indicating different configurations of reasoning and coordination"

Discussion¶

Related Claims¶

• A secondary ENA adding ICAP engagement codes showed proactive interactions linked constructive engagement with adequate reasoning, while reactive dialogues emphasised procedural pairings — possibly the same claim (merge candidate)
• Epistemic network analysis shows metacognitive behaviors co-occur with invitation actions in DR and DE but with human-initiated reasoning and elaboration in CI — related
• Condition-specific mediation estimates showed a literacy–performance association in the reactive condition but not the proactive condition, with indirect effects not significant — related
• No significant GenAI-literacy-by-agent-condition interaction emerged, so the formal models did not support different literacy slopes across designs — related
• Human-AI tutoring increases IXL time on task by 25% with no significant difference between proactive and reactive groups — related
• Human and AI discourse differ functionally: students lead invitations, connection, and regulation moves, while AI leads reasoning, coordination, and agreement — related