In Study 2, Question Autocompletion was widely adopted and the proportion of knowledge-based questions increased significantly across session stages¶

Claim · All claims
Evidence · 2 studies · 1 associational r1, 1 design r3 · q3
Currency · no dated study recorded · not yet checked against its sources · not reviewed by a person

Subclaims¶

q3 i? 74.9% of all submitted queries in Study 2 utilized autocompletion. → autocomp-adoption q3 i? The proportion of knowledge-based questions rose progressively across early, mid and late session phases (18.85%, 37.19%, 40.43%; Friedman p=0.015; early vs. late Wilcoxon p=0.019). → autocomp-scaffold

Evidence¶

autocomp-adoption¶

Jeongmin Rhee, Changhee Lee, Hyunwoo Kim, Kiroong Choe, Bohyoung Kim, Sungahn Ko, and Jinwook Seo. 2026. Penquiry: A Pen-based Interactive In-situ Q&A System Leveraging LLMs. https://arxiv.org/abs/2609.19870 doi:10.48550/arxiv.2609.19870

q3 · i? · design · r3

Study 2 usage analysis (N=16, within-subjects V1 vs. V2): the large majority of submitted queries used autocompletion; participants reported it "reduced the burden" of writing full sentences, while the retained Free Question option preserved autonomy.

"The effectiveness of the Question Autocompletion is evidenced by its high adoption rate, with74.9% of all submitted queries utilizing autocompletion."

autocomp-scaffold¶

Jeongmin Rhee, Changhee Lee, Hyunwoo Kim, Kiroong Choe, Bohyoung Kim, Sungahn Ko, and Jinwook Seo. 2026. Penquiry: A Pen-based Interactive In-situ Q&A System Leveraging LLMs. https://arxiv.org/abs/2609.19870

q3 · i? · associational · r1

Study 2 temporal analysis of question content across three chronological session phases (Figure 9), within the V2 condition; the printed Friedman and Wilcoxon statistics support a progressive increase in knowledge-based questions. No standardized effect size is printed.

"Results revealed a significant progressive increase in the proportion of knowledge-based questions:18 .85%(early),37 .19%(mid), and40 .43%(late) (Figure 9; Friedman test, 𝑝=0.015; Wilcoxon signed-rank test for early vs. late,𝑝=0.019)."

Discussion¶

Related Claims¶

• Student chatbot queries were overwhelmingly task-oriented initial inquiries, dominated by definition requests and pasted assignment questions rather than iterative critique — related
• Explanatory inquiries, factual queries, and tutorial requests are the most frequent LLM learning tasks — related
• NLP models classify factual (knowledge transmitter) learner questions well but face challenges distinguishing higher-order facilitator and co-learner questions — related