Dr. Chang has a wide range of interests in theoretical and applied statistics, including time-to-event (survival) and longitudinal data analysis, missing data (competing risks and informative dropout), causal effect modeling (propensity score and marginal structural modeling), design and analysis of observational studies and clinical trials, design and analysis of studies of biomarkers in risk prediction, dynamic prediction, and machine learning techniques. She has served as the lead statistician on numerous research projects and has been the consulting statistician for several K-award projects. In these roles, she has helped investigators throughout the University develop new research protocols and data analysis plans for a wide range of biomedical studies and has overseen the data management and analyses of these studies.

Rather than limiting her role to applying traditional statistical methodology to projects, Dr. Chang actively encourages and promotes the use of the most up-to-date appropriate statistical methods. She has applied these methods to a wide range of investigations, including research on aging, HIV/AIDS and other infectious diseases, heart diseases, liver transplantation, health services research, and acute illness.