How the Shift from Interviews to Web-Mail Self-Administration Requires Changing How We Design Questionnaires (303123)

Keywords: Self-Administration, Web-Mail, Address-Based Sampling

It seems increasingly clear that the prevalence of telephone voice surveys is declining as a means of conducting household surveys. In addition, the lack of an email sample frame combined with the excellent coverage provided by address-based sampling, results in the development of web-push methodologies, which are now capable of achieving higher response rates than either telephone or email-only surveys, with less nonresponse bias, provided paper follow-up is provided (see Dillman, Smyth, and Christian, 2014, Internet, Phone, Mail and Mixed-Mode Surveys for summary of the research leading to this conclusion). The joint reliance on mail contact plus the offering of web and mail as response modes is raising critical new issues about where and how to build the interviewer function (e.g., offering persuasive arguments for responding, explaining the nature of the surveys, and providing answers to likely questions) into survey designs. Adjusting to this new reality requires considerable rethinking about how to communicate by self-administration alone to respondents and, building necessary portions of the interviewer function into questionnaires and other communications, without biasing recipients of the request to respond. In this paper I will articulate the theoretical basis of making this transformation in survey design requirements and provide multiple examples of how aspects of the interviewer function are strategically built into various parts of the data collection process---contact letters, follow-up requests to respond, and the questionnaire itself.