Accuracy of census report questioned

Questions are being raised about the accuracy of the Census Bureau's recent current population survey (CPS), which detailed Ohio-specific data on income, poverty, and health insurance coverage. In Youngstown, local officials question the placement of their city as having one of the lowest median household incomes in the country (source: The Vindicator, Aug. 30, 2007). In Portage County, questions are also being raised about the CPS numbers (source: The Record Courier, Aug. 30, 2007).

There are several reasons people are questioning the survey results. As mentioned in the Record Courier article, "Data for the census report is based on samplings of areas with errors possibly occurring from how respondents interpret questions and the willingness of people to accurately answer questions, among other issues, according to the bureau."  The Vindicator mentions that "The bureau's estimate comes with a very large margin of error."

The reason there is such a large margin of error is that the the CPS is an interstate survey, designed to collect data across the entire country. As a result, the survey size within Ohio is limited to a few thousand respondents. This makes the CPS's Ohio results less reliable than an intrastate survey like the Ohio Family Health Survey, which has a very large sample size (40,000 households) and therefore has a low margin of error. However, a large survey like that can only be done every few years because conducting it is a time and resource intensive process. By contrast, the CPS is conducted every year, giving more current data even if it isn't as accurate as a larger survey.