Work

From Typing to Vibes: a Systematic Review of AI Programming Assistants and Natural Language Programming in Introductory Programming Courses

Year: 2026
Type: review
Study design: Systematic Review
Abstract: The integration of large language models (LLMs) into computer science education has largely abstracted manual code writing, which fundamentally changes how students approach problem solving. This technical shift poses a critical challenge, especially to introductory programming (CS1) courses. Students are now commonly utilizing generative AI tools to bypass essential programming steps, which potentially leads to degradation in their problem-solving capabilities. Therefore, this thesis aims to systematically evaluate the integration of AI code generation or natural language programming within computing education. For this purpose, a systematic literature review (SLR) that follows the PRISMA 2020 guidelines is conducted. The thesis comprehensively analyzes 29 publications from the post-GPT era (2022-2026). The findings reveal notable patterns among novices: although AI significantly enhances productivity and completion rates, most students encounter an over-reliance risk with behaviors such as monolithic prompting and passive acceptance. As a result, with the performance failing to match initial expectations, many learners experienced frustration, increased cognitive load, and skill degradation. In response to this behavioral shift, academic institutions are advised to adapt with new academic policies, assessment methods, and curriculum designs. This includes establishing acceptable use policies, using process-based evaluations, and integrating AI literacy programs into curricula.
Author: Dinh, Tri
Language: English
Open Access status: green
OpenAlex ID: W7165827823
FWCI:
Cites: 37
Cited by:
Related to: 10