Modeling Superpopulation Variance: Its Relationship to Total Survey Error

• Activity Number: 233
• Type: Contributed
• Date/Time: Tuesday, August 10, 2004 : 12:00 PM to 1:50 PM
• Sponsor: Section on Government Statistics
• Author(s): James R. Knaub, Jr.*+
• Companies: Energy Information Administration
• Keywords: regression model ; performance measure ; scatterplot edits ; inherent variance ; nonsampling error ; sampling error

Abstract

In repeated surveys there generally are auxiliary/regressor data available that are related to data collected in a current sample or census survey. These regressor data can be used to edit the current data through scatterplots, and to impute for missing data through regression. Another use for regressor data may be the study of total survey error. To do this, stratify data by regression model application. Then the related scatterplots can be used for editing, and predicted values can be found for data not collected, if any, and also to replace all data that are collected. If ratio regression is used, variance proportionate to the measure of size, then the sum of the predicted values equals the sum of the observed values they replace. The standard error of the total of the predicted values for every member of a finite population, divided by that total, and expressed as a percent, could be labeled as the estimated relative standard error of the superpopulation, or the RSESP. The RSESP would be influenced by (1) the models chosen, (2) inherent variance, and (3) total survey error (sampling and nonsampling error). This paper proposes the RSESP as a survey performance indicator.