Dies ist eine Übersichtsseite mit Metadaten zu dieser wissenschaftlichen Arbeit. Der vollständige Artikel ist beim Verlag verfügbar.
Tool or Toy? Student Views on ChatGPT in Culturally Responsive Computing Education: A Preliminary Study
0
Zitationen
8
Autoren
2025
Jahr
Abstract
Despite ongoing efforts to diversify engineering, underrepresented minority (URM) students face persistent systemic barriers to equitable participation. Beyond just access, culturally relevant pedagogy (CRP) and near-peer mentorship have shown to improve student engagement and retention. However, the currently sparse pool of URM STEM graduates limits students' access to formally educated mentors of similar cultural and ethnic backgrounds, challenging the scalability of representative CRP-based engineering programs. Free AI tools like large language models (LLMs) can offer novice programmers natural language support for debugging code and exploring new concepts, but it remains unclear whether they can provide enough technical support to help scale URM-led engineering education programs. We investigate the potential for LLMs, specifically ChatGPT 3.5, to meaningfully support CRP in an embedded systems summer course taught at a community center within the AVELA - A Vision for Engineering Literacy & Access framework. We ask how URM students will naturally adopt ChatGPT within the educational program when presented without scaffolding: as a learning tool to enhance their knowledge of class-related content, or as a toy that distracts from it? Analysis of classroom observations, student surveys, final presentations, and individual ChatGPT logs revealed a disconnect between students' personal interest and the course objectives. While students were able to use ChatGPT for course-related learning, they rarely did so. Instead, they primarily used it to enhance their knowledge of topics related to personal interest, suggesting a lack of motivation or perceived relevance of ChatGPT as a tool, rather than a lack of ability to use it. These preliminary findings suggest that without intentional integration and guidance, CRP frameworks cannot expect students to effectively leverage ChatGPT to enhance engineering related content.
Ähnliche Arbeiten
Explainable Artificial Intelligence (XAI): Concepts, taxonomies, opportunities and challenges toward responsible AI
2019 · 8.260 Zit.
Stop explaining black box machine learning models for high stakes decisions and use interpretable models instead
2019 · 8.116 Zit.
High-performance medicine: the convergence of human and artificial intelligence
2018 · 7.493 Zit.
Proceedings of the 19th International Joint Conference on Artificial Intelligence
2005 · 5.776 Zit.
Peeking Inside the Black-Box: A Survey on Explainable Artificial Intelligence (XAI)
2018 · 5.438 Zit.