Validity and Efficiency in Analyzing Ordinal Responses with Missing Observations

November 26, 2015, 2:30pm
Location: Department of Mathematics and Statistics, University of Guelph
Room: SSC 1511
Speaker: Changbao Wu, University of Waterloo

Abstract: Ordinal responses are commonly used in psychosocial and health surveys to measure performance, behaviour, attitude and opinion, severity of diseases and health conditions. Missing values are also commonly encountered for those types of surveys. This talk focuses on two critical aspects of analysis of ordinal responses in the presence of missing observations: validity and efficiency. While validity depends largely on the missing mechanism and the methods for handling missing data, efficiency gain is closely related to the type of inference problems. We discuss efficiency scenarios for two commonly encountered problems: (i) estimation of the mean responses, and (ii) generalized regression analysis, with a major interest on using fractional imputation for handling missing data. This talk is based on two papers: the joint paper with Xichen She and the joint paper with Jiwei Zhao and Richard Cook.

This talk is sponsored by the CANSSI CRT Project “Statistical Inference for Complex Surveys with Missing Observations”.

Back to Seminar Series Schedule.

Comments are closed.