How we arrived at the data.
The 2014 State of the City Poll, sponsored by Atlantic Media and Siemens, obtained telephone interviews with a nationally representative sample of 1,656 adults living in the continental United States. The survey was conducted by Princeton Survey Research Associates International. Interviews were done in English and Spanish by Braun Research Incorporated from July 23 to August 4, 2014. Statistical results are weighted to correct known demographic discrepancies. The margin of sampling error for the complete set of weighted data is ±3.4 percentage points.
A combination of landline and cellular random digit dial (RDD) samples was used to represent all adults in the continental United States who have access to either a landline or cellular telephone. Both samples were provided by Survey Sampling International, LLC according to PSRAI specifications. The samples were disproportionately-stratified to increase the incidence of respondents in urban areas. Within the urban sample alone, the margin of error is ±4.7 percentage points.
On How We Defined Urban, Suburban, and Rural: PSRAI uses telephone samples from Survey Sampling International (SSI). SSI uses Census Bureau definitions and Census tract information to code each of the 64,000 landline telephone exchanges in the country as Urban/Suburban/Rural. In short, an exchange is coded as Urban if a plurality of the directory-listed telephone households are in a principal city of a Metropolitan Statistical Area (MSA). If a plurality of the telephone numbers are in any portion of an MSA county that is not in a principal city, then the exchange is coded as Suburban. If a plurality of the exchange’s numbers are in non-MSA counties, it is coded as Rural. Thus, for each landline telephone number dialed, there is a designation of Urban, Suburban or Rural, a variable called USR. This is a variable at the telephone number level, not at the county level. For a typical national landline telephone sample, around half of the numbers are suburban, just about three out of ten are urban and under 20 percent are rural. Cell phone numbers are coded as Urban/Suburban/Rural based on respondent self-reported ZIP code.
Weighting and Analysis: Weighting is generally used in survey analysis to adjust for effects of the sample design and to compensate for patterns of nonresponse that might bias results. The weighting in this survey was accomplished in multiple stages to account for the disproportionately-stratified samples, the overlapping landline and cell sample frames, household composition, and differential non-response associated with sample demographics.