Nonresponse in Web Surveys

Abstract

This chapter is concerned with nonresponse in Web surveys. It is easier to measure rates of nonresponse than the resulting errors and, not surprisingly, most studies have focused on the former. Still, the chapter also reviews the small body of studies that investigate nonresponse error in Web surveys. In addition, it reviews research on response rates in Web surveys and their determinants (prenotification, the mode and content of the invitation, the number of contacts, and incentives). It appears that response ratesare noticeably lower in Web surveys than in other modes, suggesting that Web surveys may be at higher risk of nonresponse error than surveys using more traditional methods of data collection. Finally, Chapter 3 discusses breakoffs and item nonresponse. These are easier to measure and study in Web surveys than in other modes, since Web surveys can capture the behaviors preceding missed questions or breakoffs when they occur.