Exploring the Application of Generative AI in Educational Psychological Measurement: Virtual Personas and Their Validity (Poster 13)

• In Event: Division D Graduate Student Committee In-Progress Research Gala Poster Session

Abstract

This study explores Generative AI in educational psychometrics by generating 2,000 virtual student personas using ChatGPT to complete AMS, MSLQ, and AEQ scales. Reliability and validity were analyzed through descriptive statistics, correlation analysis, and principal component analysis (PCA), comparing virtual data with published human subject results. Findings show high consistency in scores and factor structures between virtual and human data, while Prompt design influences responses. This research highlights a low-cost, efficient method for psychometric validation and showcases ChatGPT’s potential in modeling psychological behavior, advancing the integration of educational psychology and AI technology.

Author

• Huanxiao Wang, University of Pennsylvania