The National Institute of Statistical Sciences (NISS), a nonprofit organization that fosters high-impact cross-disciplinary and cross-sector research involving statistical sciences, held a workshop for its affiliates and others on October 17 at the Bureau of Labor Statistics Conference Center in Washington, DC. More than 70 people attended the workshop, titled “Analyzing Complex Survey Data with Missing Item Values.”

The workshop focused on the current state of research and applications for work with incomplete data and imputation for complex designs, technology transfer, application context, and dominant features that affect feasibility and statistical properties. Later in the day, the group discussed prospective joint work they could conduct.

Speakers included John Eltinge, Bureau of Labor Statistics; Phil Kott, RTI International; Rod Little, University of Michigan; Shu Yang, Harvard School of Public Health; Joe Schafer, U.S. Census Bureau; Kirk White, U.S. Census Bureau; Martha Stinson, U.S. Census Bureau; Tim Keller, USDA National Agricultural Research Service; Jerry Reiter, Duke University; Jae-kwang Kim, Iowa State University; and Paul Biemer, RTI International and The Odum Institute at The University of North Carolina. There also was a working session, “Challenging Problems with Incomplete Data and Imputation in Large-Scale Federal Surveys,” with Geoffrey Paulin of the U.S. Census Bureau.

Copies of all the presentations can be found at the NISS website: https://www.niss.org/events/niss-workshop-analyzing-complex-survey-data-missing-item-values