Uso de chatbots basados en LLMS para el aprendizaje de la programación.
Acta ScientiÆ InformaticÆ · 2025
DOI: 10.21897/26192659.3902 ↗Kurz gefasst
Fachartikel in der Zeitschrift Acta ScientiÆ InformaticÆ, 2025. Verfasst von Jorge Hernandez Morelo und 2 weiteren Autor:innen. Die Arbeit nennt 20 Literaturangaben.
Abstract (Originalsprache)
This study reports the design, implementation, and evaluation of an instructional strategy using Large Language Model (LLM) chatbots to support web programming (HTML5/CSS/JS) learning in 9th‑grade students from a public school in Monteria (Colombia). We followed a three‑cycle action‑research approach with 35 convenience‑sampled students. The intervention combined standard ChatGPT and a custom GPT (tutor role, curated prompt library, QR access), adapted to a resource‑constrained setting with unstable connectivity. Instruments included a diagnostic test, post‑test, performance rubrics, and a perception survey. Results showed gains in HTML5 structure/semantics, basic styling, and debugging: the average score increased from 3.0 to 4.2/5.0; 80% of students rated the chatbot positively for immediate feedback and availability. Findings align with recent evidence, highlighting opportunities (personalized scaffolding) and risks (hallucinations, bias, digital divide). We conclude that—with teacher mediation and active verification—LLM chatbots strengthen programming learning.
Zitieren
APA
Morelo, J. H., Pérez, S. Q., & Giraldo, J. C. (2025). Uso de chatbots basados en LLMS para el aprendizaje de la programación. Acta ScientiÆ InformaticÆ. https://doi.org/10.21897/26192659.3902
Vancouver
Morelo JH, Pérez SQ, Giraldo JC. Uso de chatbots basados en LLMS para el aprendizaje de la programación. Acta ScientiÆ InformaticÆ. 2025. doi:10.21897/26192659.3902
BibTeX
@article{Morelo2025,
author = {Morelo, Jorge Hernandez and Pérez, Santiago Quintero and Giraldo, Juan Carlos},
title = {Uso de chatbots basados en LLMS para el aprendizaje de la programación},
journal = {Acta ScientiÆ InformaticÆ},
year = {2025},
doi = {10.21897/26192659.3902}
}
EUVIMED