Assessing the Impact of Missing Data on Variance Estimates for the Medical Expenditures Panel Survey: Insurance Component (MEPS-IC)

• Activity Number: 300
• Type: Contributed
• Date/Time: Tuesday, August 3, 2010 : 8:30 AM to 10:20 AM
• Sponsor: Business and Economic Statistics Section
• Author(s): Matthew Thompson*+ and Anne Kearney
• Companies: U.S. Census Bureau and U.S. Census Bureau
• Keywords: establishment survey ; variance estimation ; jackknife ; missing data

Abstract

The MEPS-IC is an annual establishment survey that collects data about insurance options made available to employees. The MEPS-IC imputes data to account for item nonresponse. If standard variance formulas are used without any adjustments, the imputation procedure likely causes an underestimate of the true variance. The goal of this research is to obtain an approximate measure of the effect of imputation on the standard variance estimates. Three different variance estimation methods will be used in this evaluation: standard jackknife estimation, jackknife estimation with reimputed replicates, and jackknife estimation with adjustments made to imputed values, as proposed by Rao and Shao (1992), which results in unbiased variance estimates. Preliminary results indicate that the baseline estimates are within 10 percent of the adjusted estimates, though issues remain with some variables.