Unpacking Vibe Coding: Help-Seeking Processes in Student-AI Interactions While Programming

2026-04-29

AI Summary

This study conceptualizes student collaboration with generative AI in programming—termed “ambient coding”—as a form of help-seeking behavior and investigates how students of varying performance levels differ in their interaction strategies and the subsequent impact on learning outcomes. Analyzing 19,418 dialogue turns from 110 undergraduate students through inductive coding and heterogeneous transition network analysis, the research finds that high-performing students predominantly engage in instrumental help-seeking (e.g., posing questions and exploring alternatives), which elicits tutorial-like responses from the AI. In contrast, low-performing students tend toward executive help-seeking, frequently requesting direct answers, resulting in insufficient cognitive engagement. This work is the first to frame ambient coding as help-seeking behavior, revealing that current AI systems passively mirror student intent without providing pedagogical guidance, and calls for the design of educational AI capable of detecting ineffective delegation and fostering inquiry-oriented interactions.

Abstract

Generative AI is reshaping higher education programming through vibe coding, where students collaborate with AI via natural language rather than writing code line-by-line. We conceptualize this practice as help-seeking, analyzing 19,418 interaction turns from 110 undergraduate students. Using inductive coding and Heterogeneous Transition Network Analysis, we examined interaction sequences to compare top- and low-performing students. Results reveal that top performers engaged in instrumental help-seeking -- inquiry and exploration -- eliciting tutor-like AI responses. In contrast, low performers relied on executive help-seeking, frequently delegating tasks and prompting the AI to assume an executor role focused on ready-made solutions. These findings indicate that currently generative AI mirrors student intent (whether productive or passive) rather than optimizing for learning. To evolve from tools to teammates, AI systems must move beyond passive compliance. We argue for pedagogically aligned design that detect unproductive delegation and adaptively steer educational interactions toward inquiry, ensuring student-AI partnerships augment rather than replace cognitive effort.

Problem

• vibe coding
• help-seeking
• generative AI
• student-AI interaction
• cognitive effort

Innovation

• help-seeking
• vibe coding
• generative AI
• Heterogeneous Transition Network Analysis
• pedagogical alignment