How You Ask Counts

Detailed analyses of response rate data from a two-stage experiment embedded in a web-based survey show that theoretically consistent Internet-related strategies for meaningfully improving response rates to web- based surveys exist and that further attempts to identify these mechanisms should focus on invitation and reminder e-mails.

Abstract

Previous research on web-based surveys lacks systematic tests of Internet-related factors of unit response rates. Strategies that do not rely on more expensive and time-consuming methods to improve response rates are important for maximizing the utility of this survey mode in social science research. This article presents detailed analyses of response rate data from a two-stage experiment embedded in a web-based survey. These data showthat contrary to some previous assertions, theoretically consistent Internet-related strategies for meaningfully improving response rates to web-based surveys exist and that further attempts to identify these mechanisms should focus on invitation and reminder e-mails. This idea is a departure from previous analyses that examine survey content and presentation for sources of variability around web-based survey response rates.