Examining Relationships Between Engagement and School Culture in Student Cheating Behaviors and AI Chatbot Usage

• In Event: Revolutionizing Instructional Technology

Abstract

This study reports an analysis of high school survey data to understand the factors associated with cheating and the use of generative AI chatbots (e.g., ChatGPT) by high school students. Using survey data, the study uses linear and logistic regression models to characterize the relationship between various independent variables, such as behavioral engagement, homework hours, academic worry, and students’ preferred allowance of AI in various tasks. The results reveal that behavioral engagement inversely affects cheating and many instances of AI use. Believing AI is allowable for specific tasks is associated with using AI for those tasks. Believing AI use is acceptable for writing full papers is associated with increased cheating.

Authors

• Victor R. Lee, Stanford University

• Rosalia Zarate, Facebook

• Denise C. Pope, Stanford University

• Sarah B. Miles, Stanford University

• Samantha Turner Selby, Challenge Success