Skip to content
Open access

Beyond Syntax and Tracing: A Multidimensional Diagnostic Framework for Novice Code Comprehension

Oct 2026 · Anais do XXXVII Simpósio Brasileiro de Informática na Educação (SBIE 2026) · 0 citations · 17 references

Abstract

Understanding code is a major challenge for beginning programming students, especially as Large Language Models (LLMs) have made code generation easier than understanding. Despite the availability of various assessment instruments, many approaches still analyze this skill through isolated tasks and results based solely on correct answers or grades. To address this problem, this work proposes the Three-Dimensional Comprehension (3DC) model, a multidimensional approach to diagnosing code comprehension. To operationalize the model, the LIFC process was developed, integrating multiple assessment instruments and tasks to collect and triangulate evidence generated by students during program reading and interpretation activities. The proposal also defines the LOAD diagnostic profiles, which represent different comprehension patterns observed among beginning students. Preliminary results indicate that integrating multiple pieces of evidence enables the identification of specific difficulties that traditional assessments based solely on final performance cannot adequately capture. Thus, the 3DC model seeks to support more detailed diagnoses and more targeted pedagogical interventions in introductory programming education.

Read PDF

We use cookies to run the site and, with your consent, for analytics and to show ads. See our Cookie Policy.