Speaker: Dr. Hui-Fang Chen

Time: 3:10-4:30 pm, 4 June, 2019

Venue: #1113, Wang Kezhen Building

Abstract: Recent development in item response theory (IRT) proposes tree-like structures to deal with Likert scale items, and it is usually denoted as item response tree (IRTree) models. This approach provides a finer responding process to describe how an individual endorses a response category. IRTree models provide estimations of a target latent ability and the tendency of choosing a specific response category (termed response styles). Based on IRTree model and empirical evidence, this study proposes a more general framework, in which the generalized graded unfolding model and the graded response model are integrated into a two-decision structure. The new model (two-stage unfolding tree model, namely UDtree) was evaluated via simulations and an empirical study of the 2016 European Social Survey with 11 items measuring the attitude toward welfare state policies. The simulation results showed that the parameters of the UDtree could be recovered very well. Furthermore, it showed that the UDtree model fitted the empirical data better than other models and can provide more insights on response processes and the measured latent trait.

Bio: Hui-Fang Chen is currently working as Assistant Professor in the Department of Social and Behavioural Sciences at the City University of Hong Kong. She is also a member of the Psychometric Lab at CityU. Her research interests and focus include scale and measurement development, validity and reliability, item response theory and Rasch analysis and their applications to educational, psychological, and medical measures. She served as a co-investigator to evaluate “the Pilot Scheme on On-site Pre-school Rehabilitation Services (OPRS),” funded by the Department of Social Welfare, Hong Kong; the program committee of Taiwan Computer-Human Interaction (TAICHI) 2017; and the editorial board of Journal of Taiwan Occupational Therapy Research and Practice. She also serves as reviewers of several international journals in psychometrics (e.g., Applied Psychological Measurement and Journal of Educational Measure) and the annual meetings of American Educational Research Association.

Host: Prof. Yiqun Gan

2019-05-30