Multiple imputation offers a general purpose framework for handling missing data, protecting confidential public use data, and adjusting for measurement errors. These issues are frequently encountered by organizations that disseminate data to others, as well as by individual researchers. Participants in this workshop will learn how multiple imputation can solve problems in these areas, and they will gain a conceptual and practical basis for applying multiple imputation in their statistical work. Topics include the pros and cons of various solutions for handling missing data; the motivation for and general idea behind multiple imputation; methods for implementing multiple imputation including multivariate modeling, conditional modeling, and machine learning based approaches; methods for checking the adequacy of imputations via graphical display and posterior predictive checks; and applications of multiple imputation for scenarios other than missing data.  The course will have an online format spread over two half days.  The first half day will cover general methodologies, and the second half day will cover applications.

Instructor Bio:
Jerry Reiter is Professor of Statistical Science at Duke University. He also serves as Chair of the Department of Statistical Science at Duke. He received a PhD in statistics from Harvard University in 1999, and a BS in mathematics from Duke University in 1992. His main research areas include methods for handling missing data, for protecting confidentiality in public use data, and for integrating information from multiple sources. He is a Fellow of the American Statistical Association, and recipient of the Gertrude M. Cox Award and the Youden Award.