Differences between nonrespondents and refusers in market surveys using mixed modes of contact

Abstract

When one is conducting survey research under mixed modes of contact, which incorporate both an interviewer-present component and a self-administered component (i.e., drop-off delivery, prenotification, “foot-in-the-door” techniques), potential respondents may choose to not participate in the study in two distinct ways. First, they may refuse or not refuse the interviewer’s request for participation. Later, they may respond or not respond to the self-administered questionnaire or other data collection form. As such, surveys conducted under such mixed modes of contact afford the researcher an opportunity to search for differences between refusers and nonrespondents in survey research. This article reports on the results of one such study incorporating a classic footin-the-door design. Of a sampling frame of 2,500 households, 2,128 were contacted and completed a short preliminary telephone interview seeking mostly demographic information. Of those, 257 (12%) refused the interviewer’s request to complete a longer mailed questionnaire and were classified as refusers. Another 857 (40%) agreed to complete the mailed questionnaire, but they did not and were classified as nonrespondents. Demographic comparisons revealed substantial differences between the two nonparticipation groups and respondents-differences described by the researchers as largely following family life cycle patterns. Nonrespondents tended to be disproportionately represented by Full-Nest I consumers (e.g., compared to respondents, 12% more nonrespondents were under 35 years old, 6% more had dependent children living at home, 10% more of those had youngest child under 6 years of age, and 7% more were employed outside the home.) On the other hand, refusers tended to be disproportionately represented by Empty-Nest II consumers and Solitary Survivors (e.g., compared to respondents, 9% more were over 50 years old, 16% fewer had dependent children at home, 11% fewer were employed, 13% more had annual income under $10,000, and 10% fewer were married).