An Examination of the Relative Variance of Replicate Weight Variance Estimators for Ratios through First Order Expansions

• Activity Number: 414
• Type: Contributed
• Date/Time: Tuesday, July 31, 2012 : 2:00 PM to 3:50 PM
• Sponsor: Section on Survey Research Methods
• Author(s): Jerry Maples*+
• Companies: U.S. Census Bureau
• Keywords: small area ; variance models ; replicate weight

Abstract

Replicate weight variance estimators are used to estimate the sampling variance for survey estimates in many complex governmental surveys such as the American Community Survey (ACS), Current Population Survey (CPS), and Survey of In- come Program Participation (SIPP). Small area models for the sampling variance require specication of the relative variance (square of the coefficient of variance, CV2, or similarly, degrees of freedom for a Chi-square distribution) for the variance estimator. In this paper, we will approximate the relative variance of the replicate weight variance estimator for ratios using rst-order Taylor expansions. The result of this work is to guide the model-based assumptions for small area variance models.