Authors:

William McMahon

Date Created:

Not specified

Course Title:

Professor:

Not specified

About Paper:

Large Language Models (LLM) are excellent tools for programmers when used intentionally. They are especially useful in helping novice programmers accelerate their learning curve, as LLM can bridge the pre- requisite knowledge gaps. LLM however are very broad in their output, and novice programmers lack the basic syntax knowledge and general programming skill to understand how to properly debug issues that could arise while coding. To solve this problem, this research aims to develop simple prompts and guidance for novice programmers in their interactions with LLM, so that they have a smoother learning experience with the LLM, thus minimizing cognitive overload and frustration. Rising Juniors and Seniors in the greater Lafayette area participated in PITCH, a two-week semiconductor focused program that culminated in a final design project. This final project consisted of 4-5 students working together with a browser based LLM to create a script for a "lilybot", a simple two wheeled robot with numerous sensors. We qualitatively analyzed the student interactions with Coderobots.ai (an online Chatbot powered through OpenAI's ChatGPT) and studied the questions asked by the students and the challenges they encountered in their learning process. Preliminary conclusions indicate that due to the lack of technical knowledge, students were unable to properly articulate their requests, thus making it difficult for the LLM to create the desired scripts. This research seeks to evaluate how structured support and guided prompting can enhance the effectiveness of LLM's in helping novice programmers generate functional python scripts. Keywords: Artificial Intelligence; Prompt Engineering; Programming Education; AI- Assisted Programing; Python † Presenting Undergrad Author; ‡ Contributing Undergrad Author; * Undergrad Acknowledgment

Source:

Purdue University / 2025

Topics:

No topics listed

Co-authors:

William McMahon