Join us for the Quantitative Psychology Colloquium with Guangjian Zhang (Associate Professor of Psychology, University of Notre Dame)
Title: Using Artificial Intelligence to Facilitate Derivation and Coding in Quantitative Methods Development: A Case Study
Abstract: Developing new statistical methods often requires complex mathematical derivation and implementation in a computer program. We explore the feasibility of leveraging rapidly advancing artificial intelligence (AI) to assist with these two tasks. We compare how effectively four AI models (ChatGPT 5, ChatGPT 4o, Gemini 2.5 Pro, and Gemini 2.5 Flash) assist in developing statistical tests that recast a challenging missing-data mechanism (MNAR: missing not at random) as a more manageable one (MAR: missing at random). We find that (1) AI models can meaningfully assist researchers with both derivation and implementation, and (2) ChatGPT 5 and Gemini 2.5 Pro perform more satisfactorily than the other two models. We discuss strengths and limitations of using AI to support derivation and coding in quantitative research and emphasize the importance of human judgment in applying AI to these tasks.
About Guangjian Zhang: Dr. Guangjian Zhang is an Associate Professor of Psychology at the University of Notre Dame, specializing in quantitative psychology. His research focuses on (1) exploratory factor analysis (EFA), structural equation modeling (SEM), and exploratory structural equation modeling (ESEM) with intensive longitudinal data; (2) EFA/SEM/ESEM with Likert (ordinal) data; (3) developing user-friendly software to broaden access to advanced statistical methods; and (4) collaborating with substantive researchers to address their questions using EFA, SEM, and ESEM. His work has appeared in Psychological Methods, Multivariate Behavioral Research, Psychometrika, Structural Equation Modeling, and the Journal of Personality. He serves on the editorial boards of Psychological Methods, Multivariate Behavioral Research, and the Journal of Personality, and is an elected member of the Society of Multivariate Experimental Psychology (SMEP).