Report 5: Evaluation of Census Ration Estimation and Synthetic Estimation

Written by:
RR88-15

Abstract

This report supplements 4 earlier reports which evaluated methods for estimating area counts using the original counts from the decennial census. Enhancements to those reports are: (A) synthetic cells (poststrata), leading to a single estimation based on only 2 "breakeven" error variance in estimating a nationwide proportion; (B) breakeven variances based on medians along with those of earlier reports based on expected values; (C) weighted combinations of the two separate artificial sets of population counts considered earlier.

Related Information


Page Last Revised - October 28, 2021