报告时间：2019年9月26日 10:00-11:00
报告地点：明德主楼1016
报告题目：Statistical Data Integration and Inference via Multilevel Regression and Poststratification
报告摘要：
The emergence of big data in population-based studies combines macro-level observations and micro-level measurements and provides unprecedent resources for population-based studies to address policy-related questions. However, such data may not be representative of the target population as convenience or volunteer samples, a form of nonprobability-based selection. Nonprobability samples become popular with the quick collection and low cost, in contract with the rapidly declining response rate and increasing cost of probability surveys, which leads to a new direction of survey research. The statistical agencies put research priorities on data integration and record linkage. The lack of theoretical foundations under new data collection methods presents challenges to traditional design-based approaches. We develop a unified framework under multilevel regression and poststratification (MRP) for data integration and inferences and handle the methodological and computational issues on big data in the combination of probability and nonprobability-based surveys. MRP combines prediction and weighting as a hybrid approach, and stabilizes small area estimation while accounting for sample selection and response mechanisms into modeling. We use simulation studies to evaluate the frequentist properties and compare with alternative methods. The proposal is demonstrated with real-life applications.
报告人简介：
Yajuan Si graduated from Renmin University of China, School of Statistics, with a bachelor degree in 2008. Currently she is a Research Assistant Professor (tenure-track) in the Institute for Social Research at University of Michigan-Ann Arbor. She received her Ph.D on Statistical Science in 2012 from Duke University. Before joining the University of Michigan in 2017, Yajuan was an assistant professor on biostatistics at the University of Wisconsin-Madison and a Postdoctoral Research Scholar in the Department of Statistics at Columbia University. Dr Si’s research lies in cutting-edge methodology development in streams of Bayesian statistics, complex survey inference, missing data imputation, causal inference, and data confidentiality protection. She serves as the Principal Investigator leading research projects funded by NSF, NIH, USDA, and etc., and provides statistical supports to collaborators across public health and social science.