City Research Online

Abstract

Face-to-face surveys are still considered the gold standard in academic survey research despite being challenging in practical implementation and therefore resource terms. This thesis aims to contribute to a better understanding of fieldwork processes and increase their efficiency. In early phases of the survey fieldwork, researchers and practitioners face a difficult challenge: from a methodological perspective, it is necessary to establish contact with each sampled unit to avoid nonresponse or sample selection biases. At the same time, reducing the number of contact attempts as much as possible is desirable to decrease fieldwork costs. A successful contact at the first attempt would be the solution to this challenge. Consequently, knowing more about how to predict first contact success would be valuable both theoretically and practically. Previous research has already identified various ‘correlates of contact’, primarily using standard statistical techniques like logistic regressions. This thesis starts by providing a review of the literature to synthesise and clarify the state of research. Then both standard statistical as well as machine learning techniques are applied to investigate how these correlates play out in three large European countries (United Kingdom, Germany, France) engaged in the European Social Survey. This provides not only a set of empirical results for the three countries, but also a methodological comparison of the different statistical techniques, which can inform future survey fieldwork in practice and show that using machine learning approaches to answer survey methods research questions is feasible. The last empirical chapter presents a prototype for a simulation approach, which successfully tailors the attributes of a contact attempt to the characteristics of a potential respondent to maximise the probability for a successful first contact.

Publication Type:	Thesis (Doctoral)
Subjects:	H Social Sciences > HM Sociology
Departments:	Doctoral Theses School of Policy & Global Affairs > Sociology & Criminology School of Policy & Global Affairs > School of Policy & Global Affairs Doctoral Theses

Export

Downloads