Do Cell Phones Affect Survey Research?

It depends.

According to government statistics, a growing number of Americans rely solely on a cell phone for their telephone service. As of the end of 2006, 12.8% of U.S. households, with 11.8% of all adults, could be reached only by cell phone. Because most telephone surveys sample only among landline telephone numbers, cell-only adults are omitted from these surveys, thus resulting in a potential bias in the sample.


Size and characteristics of the cell-only population

Cell-only adults are significantly different in many ways from those reachable on a landline. They are younger, less affluent, less likely to be married, more likely to rent their home, more urban, and more liberal on many political questions. For some of these characteristics, the differences between cell-only and landline adults are quite large. Research by U.S. Centers for Disease Control and Prevention has found that cell-only adults are more likely to binge drink, to be a regular smoker, and to lack medical insurance.;they are also more likely to describe their health status as excellent or very good, to report having been tested for HIV, or to describe themselves as obese.

Bias in landline telephone surveys because of the exclusion of cell-only adults

Despite the large differences on some characteristics between cell-only adults and those reachable on landlines, there is little evidence that surveys based on landlines are suffering from a significant bias because of the exclusion of cell-only adults from their sampling frames. Several dual-frame telephone surveys have shown that combining interviews from landline and cell phone sampling frames, and weighting each group appropriately, does not significantly change the results from those obtained from the landline sampling frame alone. Even at 11.8% of all adults, the cell-only population is still a relatively small proportion of the total population. The differences between cell-only and landline adults are not so large that the omission of the cell-only population can shift the overall results in most survey questions. This relatively reassuring situation does not necessarily hold for certain subgroups in the population where the proportion of cell-only households is considerably larger than in the public at large. Among young people, in particular, where 25%-30% are cell-only, the potential for a significant bias is much greater.


Conducting surveys on cell phones

Fortunately for survey researchers, it is feasible to conduct surveys on cell phones, although it is more difficult to do so than on landlines. The costs are much higher, especially for surveys targeting the cell-only population. There is an incorrect assumption on the part of many people that it is illegal for survey researchers to call cell phones. Federal law prohibits the calling of cell phones with the use of automatic dialing devices, which are commonly used by both survey organizations and telemarketers. But survey organizations are permitted to call cell phones if the numbers are dialed manually. It is common practice in the survey profession to offer respondents on cell phones a small amount of money to reimburse them for the costs of the incoming call. A number of other “best practices” are evolving for the conduct of surveys on cell phones.


More Information

  • “Wireless Substitution: Early Release of Estimates Based on the National Health Interview Survey, July – December 2006. (PDF)” Blumberg, Stephen J., and Julian V. Luke. May 14, 2007a.   Report by the U.S. Centers for Disease Control and Prevention.
  • “Cell Phone Sampling Summit II Statements on Accounting for Cell Phones in Telephone Survey Research in the U.S.”   Lavrakas, Paul, and Charles Shuttles. 2005.
  • “How Serious is Polling’s Cell-Only Problem?” Keeter, Scott. 2007.Pew Research Center report. June 20, 2007.