Online Program Home
My Program

Abstract Details

Activity Number: 119 - Statistical Data Editing Modernisation
Type: Topic Contributed
Date/Time: Monday, July 29, 2019 : 8:30 AM to 10:20 AM
Sponsor: Government Statistics Section
Abstract #305078 Presentation
Title: A Generalized Framework to Evaluate Imputation Strategies: Recent Developments
Author(s): Darren Gray*
Companies: Statistics Canada
Keywords: Imputation; Simulation; Data Visualization
Abstract:

For many National Statistical Organizations, imputation is the preferred treatment for item non-response. Consequently, the choice of imputation strategy can have a significant impact on resulting statistical estimates. Recently, a generalized framework to evaluate and improve imputation strategies at Statistics Canada was proposed and used to examine the choice of imputation strategies within the confines of the Canadian Census Edit and Imputation System (CANCEIS). The goal now is to develop a generalized, user-friendly tool for survey methodologists, managers and statisticians, allowing them to assess and compare imputation strategies on existing datasets. The focus of this paper is on the development of the tool itself, in particular the choice of simulation parameters and output measures. Finally, we explore how best to present results, including data visualization, to facilitate data-driven decision-making in survey design.


Authors who are presenting talks have a * after their name.

Back to the full JSM 2019 program