Using Prediction-Oriented Software for Survey Estimation - Part II: Ratios of Totals

by James R. Knaub, Jr.

Abstract:This article is an extension of Knaub (1999), "Using Prediction-Oriented Software for SurveyEstimation," which dealt with the estimation of totals and subtotals and the corresponding estimates ofvariance in the presence of 'missing data,' whether missing as part of a sampling scheme, or as a result ofnonresponse. The current article deals with ratios of totals. An example from the electric power industrywould be the estimation of revenue per kilowatthour and its associated variance estimate. As in Knaub(1999), the goal is to produce such estimates by making use of currently available software in which themodel can be quickly and easily modified, and the data may be stored in such a manner that they may beeasily recategorized for purposes of publishing various aggregations of the data with correspondingvariance estimates.

Key Words: survey sampling, prediction, estimation, imputation, variance estimation, ratios of totals

James R. Knaub,

Editor:John Hinde,

READING THE ARTICLE: You can read the article in portable document (.pdf) format (55008 bytes.)

There was an addedum submitted in August 2012: Addendum

NOTE: The content of this article is the intellectual property of the authors, who retains all rights to future publication.

This page has been accessed 3413 times since July 24, 2006.

Return to the InterStat Home Page.