Psychological science increasingly relies on complex measurement models, large data structures, and multi-site collaborations-yet many traditional methodological tools face limitations in scalability, validity, and data accessibility. This presentation highlights how artificial intelligence (AI) can meaningfully expand the methodological toolkit available to psychological researchers. I first introduce a deep generative adversarial algorithm for high-dimensional item factor analysis that improves latent variable recovery in challenging measurement conditions. I then present a federated item response theory framework that allows researchers to estimate psychometric models across institutions without sharing individual-level data, addressing privacy concerns central to contemporary psychological research. Next, I discuss new evaluations of large language models for automated evidence synthesis and for conducting statistical power analyses, illustrating both their potential and their current methodological constraints. Finally, I examine recent work comparing structural equation modeling with targeted maximum likelihood estimation, showing how AI-driven estimation strategies can clarify causal estimands and strengthen inference. Together, these projects demonstrate how AI can enhance the precision, interpretability, and reproducibility of psychological methods.
Dr. Feng Ji is an Assistant Professor in the tenure stream in the Department of Applied Psychology and Human Development and holds a Canada Research Chair in Psychometrics and Responsible AI. He received his PhD in Biostatistics from the University of California, Berkeley, and has extensive expertise in applying, evaluating, and developing quantitative and machine learning methods for research in the behavioral, educational, and social sciences. His peer-reviewed work appears in leading methodological outlets such as Psychometrika, Psychological Methods, Journal of Educational and Behavioral Statistics, and Multivariate Behavioral Research, as well as substantive journals including Child Development and Applied Linguistics. Dr. Ji's academic scholarship is complemented by significant industry experience, having served as a research data scientist at Google and a research psychometrician at the American College Testing (ACT). He also contributes to the field as an editorial board member for several journals, including Psychological Review, Behavior Research Methods, and Child Development.
Recent Comments