Yesterday, the Centers for Disease Control and Prevention released a report detailing estimates of wireless-only households and adults by state.  For researchers this information is a valuable resource because it is the only known estimate by state of wireless-only households.   I was surprised to learn that one in four adults (25.1%) in Oklahoma are in a wireless-only household while in Vice President Biden’s home state just four percent of adults live in a cell-phone only household.  To find out the percentage in your state check out the report.

The results come from the 2007 National Health Interview Survey and the 2008 Current Population Survey’s Annual and Social Economic Supplement.   The report notes that “more than one out of every six American homes (17.5%) had only wireless telephones during the first half of 2008,” this is up from 14.7% in 2007.   Be sure to read Mark Blumenthal’s description of the methodology of the surveys used to provide the estimates.

Wireless-only households are a conundrum for researchers who conduct telephone surveys, because many pollsters use random-digit-dialing sample.  RDD sample includes listed and unlisted phone numbers while at the same time excluding many business numbers, but it does not include cell phones. 

There are several problems researchers encounter when conducting surveys on cell phones; they are more expensive because researchers are not permitted to use predictive dialers to make the call, there is the problem of determining the respondent’s home location, and the respondent must be a safe environment to take the call/conduct the survey. 

At the national and statewide level completing surveys with cell-phone only households is not too challenging because the geography is large, but becomes much more difficult when conducting surveys in smaller geographic regions including Congressional districts, counties, legislative districts and municipalities.

More and more pollsters are supplementing their landline surveys with cell phone interviewing as well.