Imputation Methods for Wave Nonresponse in Panel Surveys

• Activity Number: 483
• Type: Contributed
• Date/Time: Wednesday, August 4, 2010 : 8:30 AM to 10:20 AM
• Sponsor: Section on Survey Research Methods
• Author(s): Kyu-Seong Kim*+
• Companies: University of Seoul
• Keywords: carry-over imputation ; longitudinal regression imputation ; nearest neighborhood imputation ; row-column imputation

Abstract

We consider imputation methods for dealing with wave nonresponse in panel surveys. We have a good advantage in panel surveys unlike cross-sectional surveys that the previous data can be used to compensate for wave non-response. Some methods are presented and compared numerically through a simulation study based on Korean Welfare Panel data. Simulation results show that ratio and row-column imputation methods are much more effective in both predictive and estimation accuracy. Regression, longitudinal regression and carry-over imputation methods perform better in predictive accuracy, but nearest neighborhood, nearest neighbor regression and hot-deck imputation perform better in estimation accuracy. Finally, the mean imputation shows much lower performance in both criteria.