Student AI Use in Programming: Between Exploration and Delegation
A new arXiv study highlights how students use generative AI in programming, identifying differences in interaction patterns between high- and low-performing learners.
What happened?
Researchers have analysed 19,418 interactions from 110 university students using generative AI for programming, a phenomenon referred to as "vibe coding". The study, published on arXiv, classifies student interactions as "help-seeking" processes. It examined differences in interaction sequences between high- and low-performing students to understand how AI is used in the learning process during programming sessions.
Key facts
| Antal studenter | 110 |
|---|---|
| Antal interaktioner analyserade | 19 418 |
| Publiceringsdatum | 2026-04-27 |
| Metodik | Induktiv kodning, Heterogeneous Transition Network Analysis |
”Generativ AI omformar programmering i högre utbildning genom vibe coding, där studenter samarbetar med AI via naturligt språk snarare än att skriva kod rad för rad. Vi konceptualiserar denna praktik som hjälpsökande.”
”Resultaten visar att högpresterande studenter engagerade sig i instrumentellt hjälpsökande – förfrågan och utforskning – vilket framkallade tutorliknande AI-svar. Däremot förlitade sig lågpresterande studenter på exekutivt hjälpsökande, ofta delegerande uppgifter och uppmanande A”
”Dessa fynd indikerar att nuvarande generativ AI speglar studentens avsikt (vare sig den är produktiv eller passiv) snarare än att optimera för lärande.”
Why it matters
The study reveals that high-performing students engage in "instrumental help-seeking" through enquiries and exploration, which elicits tutor-like AI responses. Conversely, low-performing students rely on "executive help-seeking", where they delegate tasks and ask the AI to deliver finished solutions. This indicates that current generative AI mirrors the student's intent rather than actively optimising the learning experience.
Who is affected?
Students and educators in higher education are directly affected by the role of AI in programming instruction. Developers of AI tools for the education sector are also concerned, as the results highlight a demand for AI systems that can function as active "collaborators" rather than merely "tools".
What else you should know
The study utilised inductive coding and Heterogeneous Transition Network Analysis to examine the interaction sequences between student and AI. The results suggest a distinct difference in how various student groups leverage AI support.
Quick answers about this story
Vad har hänt?
När hände det?
Varför spelar det roll?
Vilka skillnader identifierades mellan studentgrupperna?
The link opens in a new window and leads to the publisher's own site.
Källan har spårats automatiskt från utgivaren via Aheadlines signalkedja.
Get similar news straight to your inbox
The reader's room
Send in a question or an addition. The newsroom reads everything before it's published and replies when relevant. No AI-generated text – just people.
Sign in to submit a comment or question.
Read the article through your role
- Decide whether this affects strategy over 6–12 months or is just noise.
- Discuss with leadership: do we own the right question or does ownership need to move?
- Ask: what risk are we taking by NOT acting on this this quarter?
Generated angle — not editorial analysis of "Student AI Use in Programming: Between Exploration and Deleg"