The American Community Survey is an ongoing survey designed to collect detailed data about the nation's population and housing. Beginning with the 2010 Census, the American Community Survey has replaced the decennial long form, allowing annual updates of information that the U.S. Census Bureau previously did only once every ten years.

We designed the American Community Survey 2012 Content Reinterview Survey to evaluate the reliability of responses to American Community Survey questions. Along with other measures of data quality such as item nonresponse and response bias, item-level measures of response reliability are important to both data users nd American Community Survey planners. Data users need to understand how errors in the data may affect the conclusions they draw from analyzing the data. American Community Survey planners need information about current data quality to develop and test methods for improving data quality in the future.

The American Community Survey 2012 Content Reinterview Survey used a simple response variance type reinterview, meaning that we asked essentially the same questions in the reinterview that we asked in the original American Community Survey interview. We selected a sample of 72,000 households from American Community Survey returns processed between January 3, 2012 and December 3, 2012. The Content Reinterview Survey sample included returns from all three data collection modes used in 2012: Mail, Computer Assisted Telephone Interviewing, and Computer Assisted Personal Interviewing. The Content Reinterview Survey was conducted only in the Computer Assisted Telephone Interviewing mode. The overall response rate for the Content Reinterview Survey was about 67 percent.