Experimental Design for the 2006 American Community Survey Content Test

The 2006 American Community Survey (ACS) Content Test is a large household sample survey (over 60,000) designed to test proposed changes to the ACS questionnaire for 2008. The original base sample design is a stratified primary sample unit design that builds on the sample design of an earlier survey, the Census 2000 Supplementary Survey. In response to changing field requirements, several changes were implemented to address total workload size and other operational issues. Additional stratification and sampling for the personal follow-up phase of the survey was added late in the design phase to help reduce the total workload. A balanced splitpanel assignment was designed to help minimize the number of field representatives (FR’s) necessary for the survey in the less densely populated areas avoiding duplication of geographic assignment areas for experimental and control panel FR’s.