• Presentation

The Role of Landline and Cell Phone Usage Patterns in Nonresponse Error Potential Among Young Adults in Telephone Surveys

Citation

Currivan, D. B., Levine, B., Mayo, N. D., & Hampton, J. C. (2010, May). The Role of Landline and Cell Phone Usage Patterns in Nonresponse Error Potential Among Young Adults in Telephone Surveys. Presented at AAPOR 2010, .

Abstract

A current challenge in conducting telephone surveys of adults is obtaining satisfactory representation among young adults. Surveys using traditional random-digit dial (RDD) sample frames of landline numbers exclude the 33 percent of adults age 18 to 24 who live in households without landline phone service. In addition to this well-documented coverage issue, a further problem is the potential difficulty in contacting and interviewing an additional set of about 19 percent of young adults who live in households with a landline phone, but primarily use a cell phone. RDD telephone surveys may further underrepresent the full population of young adults by excluding those who primarily rely on cell phones and are difficult to reach by landline phones. Young adults who primarily use cell phones may have important similarities to those who only have wireless phone service in terms of key survey measures.

To better understand the role of phone usage patterns in the potential for nonresponse error among young adults, this paper investigates the implications of patterns of landline and cell phone use among young adults for key survey outcomes. The data we use are drawn from a statewide RDD-based survey on health behaviors that targeted young adults age 18 to 24 and included both landline and cell phone numbers. Based on respondents’ answers to questions on the presence, number, and sharing of cell phones in their household, we coded each respondent into one of four categories: landline only, primarily landline, primarily cell phone, or cell phone only. We then examined patterns in both demographic characteristics and health indicators across these four categories. This analysis provided data on how young adults’ phone usage could potentially contribute to nonresponse error, beyond coverage error. We discuss the implications of these findings for the representation of young adults in RDD surveys on health behaviors.