Conferences

In this research we examine the sequences of all contacts attempts that interviewers conduct on households to obtain contact and finally cooperation with the target person. Each contact attempt within a sequence is meaningful with respect to the result of the survey request. It’s likely that different survey and fieldwork specific characteristics determine the structure of typical contact sequences to a high degree. E.g., countries which sample households rather than individuals are liable to the “gatekeeper” phenomenon, or certain sampling frames tend to lead to many ineligible cases. Where interviewers are paid per completed interview, the number of appointments is smaller. In addition, survey culture in each country plays a role in the interaction process. That is, common practices to contact respondents and, for potential respondents, manners to react to such requests prevail. The aim of this research is therefore to understand the effects of certain survey specificities on contact sequences rather than to measure the chance of success for an interview.

We use the contact forms of the ESS 2004 and 2006. One focus is to find a suitable measure of similarity of contact sequences. Optimal matching analysis is a technique which allows computing a matrix of dissimilarities between each sequence. Contrary to other classical longitudinal methods this perspective takes into account that actors engaged in the process have knowledge about “usual” contact sequences and anticipate them.

A matrix of dissimilarities converted by factorisation can map the distribution of the different countries according to their survey specificities. Clustering produces typical sequences for each country and can show the link between the countries fixed survey issues and the interviewers strategies in the fields. Such a map can help to reveal the effective specificities of the fieldwork in different context.