Students report code understanding, debugging, and AI-limitation awareness as primary learning benefits of buggy GenAI programming tasks¶
Claim · All claims
Evidence · 1 study · qualitativer3·q3
Currency · no dated study recorded · not yet checked against its sources · not reviewed by a person
Subclaims¶
Reading the evidence codes
q is the design tier, i the impact, n the sample, and the kind with its r
the rigour judged within that kind: what each code means.
This claim: no standardized effect is recorded in observations/ for it yet, so no benchmark reading is possible.
q3 i? Students framed the activity as learning by diagnosing and recovering from failure, reporting practice in reading, tracing, and repairing plausible programs. → rq3-reflections
q3 i? Roughly two thirds of students agreed or strongly agreed (~64%) that they understand self-written code better than AI-generated code. → rq3-reflections
Evidence¶
rq3-reflections¶
Victor-Alexandru Pădurean, Kaitlin Riegel, Alkis Gotovos, Jyotika Mahapatra, Ahana Ghosh, Paul Denny, Juho Leinonen, James Prather, and Adish Singla. 2026. When AI Is Wrong on Purpose: How Students Respond to Buggy GenAI Code. Proceedings of the ACM Conference on International Computing Education Research Vol. 1 (ICER 2026 Vol. 1). https://doi.org/10.1145/3765964.3811667
q3 · i? · qualitative · r3
Thematic analysis of 204 reconciled free-text reflections from the post-tasks questionnaire, with mean Krippendorff's α = 0.82. Code understanding (99 labels) and Debugging (82 labels) were the most frequent themes.
"Students often framed the activity as learning by diagnos- ing and recovering from failure as opposed to simply producing a solution."
Discussion¶
Related Claims¶
- Erroneous examples improve conceptual understanding by forcing comparison with correct models. — the general claim this finding bears on, with the studies behind it
- Injected bugs prompt more direct code edits while natural bugs prompt more reprompting in prompt-based programming tasks — related
- Students make smaller code changes and fewer edit-and-run attempts after injected bugs than after natural bugs — related
- Prompting strategies differ by bug source: natural bugs elicit specification-oriented strategies while injected bugs elicit guidance-oriented strategies — related
- Generative AI generates responses based on probability and pattern prediction, not reasoning or understanding — related
- CS1 students tend to agree it is easier to solve problems by writing natural language prompts than by writing code (mean 3.80 on a five-point scale) — related
- Students most frequently reported that prompt-based tasks were Easier/Simpler/Faster and supported analyzing and solving problems, in thematic analysis of 200 reflections — related
- Student dialogue prompted educators to reconstruct teaching beliefs about assessment, debugging instruction, and active learning — related
Page metadata
| Field | Value |
|---|---|
type |
claim |
id |
student-reflections-buggy-genai-learning-benefits |
title |
Students report code understanding, debugging, and AI-limitation awareness as primary learning benefits of buggy GenAI programming tasks |
description |
Students report code understanding, debugging, and AI-limitation awareness as primary learning benefits of buggy GenAI programming tasks |
status |
review |
evidence_strength |
moderate |
generated |
by process:wiki-ingest on 2026-10-10 |
trust |
unverified — no one has confirmed this page's content |
sources |
1 — rq3-reflections |
Frontmatter is the page's machine-readable record — see the Schema & Guide.