Document Type

Article

Publication Date

Spring 2025

Keywords

artificial intelligence, student engagement, lesson planning, large language model, educational technology, correlational analysis, digital literacy

Abstract

Artificial intelligence (AI) is a technology that continues to become increasingly more prevalent, for better or worse, in the classroom. One such way that it has seen use in the classroom is in the generation of lesson plans by teachers who use AI-driven large language models (LLMs) such as OpenAI’s ChatGPT and Google’s Gemini. To see if these LLMs are, or could be, a reliable way to craft lessons as the technology continues to improve, the purpose of this study is to determine if a correlation exists between the strength of these different LLMs (measured in “parameters”) and the levels of student engagement reported by students who engage in the lessons generated by these LLMs. The participants involved in this study were ten students across two classrooms in an alternative-setting high school located in a suburban town in Illinois. This data was then collected by having four different AI-driven LLMs with different numbers of parameters generate two lessons each, and by having the student participants fill out a seven-item engagement survey after each lesson, giving them a total average engagement score to report. A correlational analysis between the number of parameters of the LLMs and the engagement scores reported for the respective lessons was then run through SPSS to determine if a correlation existed between these two data sets.

Comments

This paper was completed as part of TCH 482, Professional Research II, and was presented at the University Research Symposium.

Download

Share

COinS