Grantee: Hui Xie
- University of Illinois at Chicago
Dr. Xie is passionate about developing proper data analytic methods to improve the scientific rigor of testing health behavior theories. Despite the best efforts of researchers, complications can happen that affect data quality and usability through selection bias in the process of measurement, sampling and data analysis. These complications can lead study findings astray if not accounted for properly. Dr. Xie is developing statistical methods to assess and address the impact of selection bias, and to improve the reliability, validity and quality of study findings in a wide variety of contexts. Examples of applications include reliable evaluation of the change of smoking behaviors over time in a cohort study of US young adults, more powerful detection of risk factors for disease occurrence, and examination of the behavioral intervention effects on smoking abstinence by accounting for participants' noncompliance behaviors, such as attrition and nonresponse.
Dr. Xie has strong interest and experiences in developing cutting-edge analytical methods for use with precise and longitudinally intensive consumer behavioral data captured in real time using electronic devices. Dr. Xie's current research project extends to developing appropriate analytical techniques and computational tools for use with new real-time data capturing methods, such as Ecological Momentary Assessment (EMA). Working with prominent researchers and colleagues, Dr. Xie focuses on Ecological Momentary Assessment of adolescent smoking behaviors with data collected using mobile devices. Despite many advantages of EMA to advance our understanding of health behaviors and their mechanisms of change, the issue of missing data in traditional studies remains and can present an even greater challenge, due to intensive measurement processes of the EMA data. One aim of the project is to develop computationally feasible methods and tools to address the issue of missing data. Because of the commonality of the issue, the methodology and software to be developed can be applied to a variety of cancer-relevant research areas, including pain, diet, exercise, and other types of intensive data in E-Health and mobile health.
I am fascinated by the capabilities of new real-time data-capturing methods to improve behavioral research. I am excited to be developing appropriate analytic techniques and computational tools for use with these new kinds of data and emerging research approaches.”