Skill-targeted explanation styles: pragmatic versus contingent LLM error-message rewrites

Design · All designs
Evidence · 2 claims (2 for) · 1 study (1 causal), q2 · 0 of 1 report an effect size · 2 claims rest on one study

Description

The article defines two skill-targeted explanation styles for LLM rewrites of Python error messages, designed for different assessed proficiency levels. As the article states, "Pragmatic messages are concise and task-oriented: they name the error, localize the likely cause, and provide a short hint toward repair", while contingent messages "provide more scaffolded guidance" including context about likely intent, possible misconception, and reasoning path. The styles are informed by actionable-feedback work, Bloom's taxonomy, claim-evidence-reasoning, and Toulmin's model, and were evaluated against standard interpreter messages in a controlled crowdsourced experiment.

Design Implications

Context

Requirements

• LLM-generated rewrites require the code snippet, the standard interpreter message, and line-numbered context in the prompt

Constraints

• The system does not dynamically adapt during interaction; the authors explicitly do not claim dynamic adaptation, only styles designed for different assessed proficiency levels

Target Learners

• novice and more experienced Python programmers in crowdsourced debugging tasks

Learning Goals

• interpreting programming error messages and repairing buggy code

Claims

• Llm Error Rewrites Subjective Not Objective Gains [+M]
• Llm Rewrites Lower Perceived Cognitive Load [+M]

Related Designs

Examples

-

Key Sources

• Alexandru-Radu Moraru, Shreyan Biswas, Ujwal Gadiraju. (2026). Beyond the Traceback: Using LLMs for Adaptive Explanations of Programming Errors. https://arxiv.org/abs/2608.20896

Page metadata

Field	Value
type	design
id	skill-targeted-explanation-styles-pragmatic-contingent
title	Skill-targeted explanation styles: pragmatic versus contingent LLM error-message rewrites
description	The article defines two skill-targeted explanation styles for LLM rewrites of Python error messages, designed for different assessed proficiency levels.
status	review
generated	by process:settle-candidates on 2026-10-11
trust	unverified — no one has confirmed this page's content
sources	1 — alexandru-radu-moraru-2026

Frontmatter is the page's machine-readable record — see the Schema & Guide.