Exclusion of cell-phone users in surveys studied
PULLMAN – It is estimated that U.S. cell-phone-only households have increased from about 4 percent in 2004 to more than 14 percent in 2007, with 33 percent of cell-phone respondents having no landline telephone, according to research featured in the spring 2009 WSU Academic Showcase.
RDD is a method for selecting people for telephone surveys by generating telephone numbers at random. It includes unlisted telephone numbers but not cell-phone numbers.
In their work at the SESRC, Tarnai, Moore and Schultz conduct phone, mail and web surveys for a variety of different agencies. They participate in conferences and are active in organizations that are always looking to improve methods used in survey research. So, Schultz said, their project was inspired by the need to stay current with what is happening in their field.
“This is what we do; this is our job,” Schultz said. “But it is more than a job; this is our passion. To learn how to improve survey research and provide valuable information to others is what drives us.”
The study aimed to determine whether or not surveys can continue to use RDD without considering the cell-phone-only population. This led to the bigger question of whether ignoring cell-phone-only users would bias the results of a survey.
The paper concluded that survey researchers must include cell-phone-only households in order to maximize survey coverage and accuracy.
Excluding cell-phone-only users excludes a particular segment of the survey-sample population. Those who only use a cell phone tend to be young, male, single and are more likely to rent a home rather than own one, the WSU researchers found. The results also showed that cell-phone-only respondents have either more or less education than people from the RDD samples.
A full version of the study, “Characteristics of Cell Phone Only, Listed and Unlisted Telephone Households,” can be found here.