American Community Survey Sample Size Research

• Activity Number: 20
• Type: Topic Contributed
• Date/Time: Sunday, August 1, 2010 : 2:00 PM to 3:50 PM
• Sponsor: Section on Survey Research Methods
• Author(s): Steven Hefter*+
• Companies: U.S. Census Bureau
• Keywords: American Community Survey ; Sampling ; Reliability

Abstract

As a replacement for the once-a-decade decennial long form sample estimates, the reliability of the American Community Survey (ACS) estimates is an important consideration. This paper presents research that assesses the reliability of the ACS estimates. The reliability of theoretical tract level estimates from the Census 2000 Long Form sample is used to determine the annual sample size and sampling rate necessary for the ACS 2011-2015 five-year estimates to achieve various levels of reliability. In this approach, the levels of reliability for the ACS are described as a function of the Census 2000 Long Form reliability, as measured by the coefficient of variation for a fixed 10 percent population estimate of the poverty rate. Note that the poverty rate estimate has been identified as a key measure for the ACS and has been shown, in general, to have relatively large standard errors.