Research Summary: Perceived usefulness and intention to use large language model-generated feedback across three educational levels: a user-centred study in programming

A user-centred study investigated the perceived usefulness and intention to use feedback generated by Large Language Models (LLMs) in programming education across three educational levels. The research evaluated LLM-produced formative feedback using a two-layer instrument, assessing both individual response feedback and consolidated performance reports on dimensions such as clarity, accuracy, actionability, overall satisfaction, and intention to use. This exploration aims to understand student perspectives on AI-generated feedback in an educational context.

Why it matters

The increasing use of Generative AI in education necessitates a robust understanding of its effectiveness and acceptance by end-users. Insights into student perception of LLM-generated feedback are crucial for guiding the responsible development and implementation of AI tools that genuinely support learning outcomes and user engagement.

Key insights

• The study focused on student perception and intention to use LLM-generated formative feedback in programming education.
• Feedback was produced by an automated, rubric-based LLM assessment.
• A two-layer evaluation instrument was used: response-level feedback (rated on clarity, specificity, accuracy, actionability, usefulness) and a consolidated performance report (rated on overall satisfaction, relevance, personalization, cognitive load, intention to use, motivation).
• The study involved 144 students across secondary education and short-cycle higher education.
• The research sought to determine if students find such feedback useful, actionable, and worth using.

Source

Frontiers in Education — https://www.frontiersin.org/articles/10.3389/feduc.2026.1934069