The hunt for the last respondent : nonresponse in sample surveys

Abstract

The Hunt for the Last Respondent has been inspired by concerns about the possibly detrimental effect of nonresponse on the accuracy of survey outcomes, as response rates are generally considered to be the most important criterion of survey quality, and the Netherlands is notorious for its low response rates. It addresses the following questions:

Why are high nonresponse rates a reason for concern? Who are less likely to respond, either because they are more difficult to contact or because they are more reluctant to cooperate? How can response rates be enhanced?

These questions are addressed in an extensive study of the international literature on nonresponse presented in part 1 of this thesis. Nonresponse is worrisome when those persons who are not contacted or who refuse to cooperate differ from respondents and these differences cannot be adjusted for by using auxiliary information, generally background variables. Part 1 gives an overview of background variables as correlates of contactability and cooperation. In addition, attention is paid to causal factors of survey participation, such as the attitude towards surveys, the topic of the survey, social isolation and social participation, and to the distinction between situational and survey-related reasons for refusal. Finally, it presents an overview of response enhancing strategies. The second, empirical part of the study elaborates on the literature presented in part 1 and addresses the questions:

How to study nonresponse? Do enhanced response rates improve the accuracy of survey outcomes? How to combat nonresponse error and allocate funds effectively?

To answer the first question, two instruments and two models are introduced. The instruments are standardized contact forms on which to register fieldwork details and a follow-up survey among persistent refusers. The first can provide information on the degree of contactability and cooperation of the respondents, the second on final nonrespondents. The models assume different relationships between difficult respondents and final nonrespondents. Nonresponse on two surveys is analysed, namely the Dutch Amenities and Services Utilisation Survey 1999 and its follow-up survey among persistent refusers, and the European Social Survey 2002/2003. In these surveys high response rates were obtained, even among adamant refusers. Based on the analyses the final two questions of this study were answered. It turns out, as others have found earlier, that enhancing response rates does not always improve the accuracy of survey outcomes. The study recommends to spend a part of the funds for data collection on obtaining information about final nonrespondents as this is more effective than raising the response rate by a few percent.