Gongjun Xu, University of Michigan

Early Career Award: Identifiability of Cognitive Diagnosis Models

Keynote Speaker

2023 Early Career Award

Cognitive Diagnosis Models (CDMs) are popular statistical tools for developing diagnostic-based assessments in education, psychology, and other social and behavioral sciences. CDMs can be viewed as a family of restricted discrete latent variable models, where the model parameters are restricted via the Q-matrix to reflect pre-specified diagnostic assumptions. Though widely used, CDMs often suffer from nonidentifiability due to the models’ discrete nature and complex restricted structure. This talk will introduce some recent identifiability results on CDMs by considering both strict and partial identifiability of the model parameters. The developed identifiability conditions only depend on the design Q-matrix and are easily checkable, which provides useful practical guidelines for designing statistically valid diagnostic tests.

About the Speaker

Dr. Gongjun Xu is an Associate Professor in the Department of Statistics with a joint appointment in the Department of Psychology at the University of Michigan. He received his B.S. in Statistics from the University of Science and Technology of China in 2008, and his Ph.D. in Statistics from Columbia University in 2013. His research interests include psychometrics, latent variable models, cognitive diagnosis modeling, item response theory, and statistical learning and inference. Dr. Xu is currently serving as Co-Editor-in-Chief for Journal of Educational and Behavioral Statistics, and Associate Editor for Journal of American Statistical Association, Annals of Applied Statistics, Statistica Sinica, and Journal of Data Science. He received NSF CAREER Award (2019), International Chinese Statistical Association (ICSA) Outstanding Young Researcher Award (2019), Bernoulli Society New Researcher Award (2019), Committee of Presidents of Statistical Societies (COPSS) Emerging Leader Award (2023), and Psychometric Society Early Career Award (2023).

Log in