Student reflections identify speed and scaffolding as the dominant theme, with verification burden, overreliance worries, and prompt precision also prevalent¶

Claim · All claims
Evidence · 1 study · qualitative r2 · q2
Currency · no dated study recorded · checked against its sources 2026-10-20 (automated, 1 of 1 entry) · not reviewed by a person

Subclaims¶

q2 i? In LLM-assisted thematic analysis of 1,158 reflections, Speed and scaffolding was the dominant theme (84.5% of students), followed by Task fit (64.5%), Verification burden (57.3%), Learning and agency (57.2%), Code quality (54.5%), and Prompt precision (54.0%). → rq3-themes

Evidence¶

rq3-themes¶

Nasser Giacaman, Valerio Terragni, Paul Denny, and Viraj Kumar. 2026. Commenting with Copilot: A Taxonomy and Multi-Year Analysis of Student Code-Generation Specifications. Proceedings of the 2nd ACM Virtual Global Computing Education Conference (SIGCSE Virtual 2026). https://doi.org/10.1145/3795867.3830978

q2 · i? · qualitative · r2

LLM-assisted four-round thematic analysis of end-of-activity reflections (Table 2), with 13 themes; Speed and scaffolding reached 979/1158 students, with Task fit at 747/1158 and Verification burden at 663/1158. No effect size is printed.

"Speed and scaffolding.This was the dominant theme. Students often described comments as a quick way to turn ideas into code, especially when getting started or avoiding low-level syntax:"

Discussion¶

Related Claims¶

• LLM-assisted inductive qualitative coding carries risks of superficial themes, broad or redundant codes, and hallucinated interpretations, so LLMs should augment rather than replace human researchers — a broader claim this one bears on
• Reviewed literature reports risks of overreliance, AI errors in complex tasks, and students' uncritical acceptance of AI-generated outputs — related
• Overreliance on AI is the most prevalent pedagogical limitation (27 studies, 50%), followed by reduced critical thinking (25 studies, 46%) — related
• Low comment rewriting does not imply a smooth process: students describe the main effort as verifying generated code rather than revising prompts — related
• Students' LLM use in CS2 fell into three themes: understanding, assisting code writing, and writing code, with only a few using LLMs to write code — related
• Students predominantly recommended future students use office hours and formal course staff, with LLMs positioned as a supplement — related
• GenAI use involves substantial hidden labor of prompting, verification, and correction that remains invisible in productivity evaluation — related
• Coded prompt logs show different prompting strategies foreground different pedagogical themes, with technical prompts the largest strategy total (398 instances) — related
• Teachers experienced AI as expanding design possibilities while increasing the burden of verification and professional responsibility — related
• Student GenAI use patterns shifted across assessments: Learning and Formatting dominated under purpose-specific use, while Feedback became the dominant use (66.7%) under open use — related
• More experienced AI users showed a Boilerplate Blindspot: uncritically accepting 'trivial' AI-generated code in which later defects originated — related
• Pedagogically scaffolded generative AI in design thinking education augments rather than replaces human judgment — related
• Students showed bounded trust and uneven verification behavior: 23 of 837 student turns (about 2.7%) contained explicit risk-aware moves — related
• An efficiency paradox: well-intentioned students default to passive GenAI processes and experience cognitive laziness and overreliance — related
• Students report awareness of feedback-acting strategies but experience difficulty carrying them out — related
• Students most often report hybrid interaction modes with generative AI, with passive task delegation and verification-only use as smaller patterns
• AI functioned primarily as a source of preliminary ideas and cognitive support rather than as a substitute for professional judgment, with students identifying contextual limitations of AI recommendations — related
• Qualitative reflections identify developmental inappropriateness (45%), vagueness and repetition (32%), and implementation feasibility (23%) as the main AI limitations — related
• The predominant student evaluation of LLMs (42%) is positive but conditional on the student's own verification work — related