Conferences

Problem statement: Orally administered interviews conducted in Swiss German speaking regions of Switzerland often diverge from the written German questionnaire (there is no written form of the many Swiss German dialects), and this may compromise standardization, measurement equivalence, and data quality. Until now, there has been no systematic evaluation of the nature and effects of such dialectal adaptation on the data collected in these Swiss German speaking regions of Switzerland.

Purpose: In June and July 2005, a study was conducted in Germany and Switzerland on oral adaptation in telephone surveys. For the current study, we follow up on this previous work by focusing on yet unanalyzed Swiss interviews (based on the German text) to assess the extent to which, the ways in which, and the reasons for which Swiss German speaking interviewers deviate from the scripted German questionnaire at various linguistic levels (i.e., phonetic, morphological, syntactic, discourse). The study focuses on systematic patterns of adaptation and their influence on data quality. The aim is to determine if consistent “corrections” (i.e., patterns of adaptation) by interviewers provide lessons for survey question development and/or interviewer training in such linguistic contexts.

Approach: Twenty five interviews from the European Social Survey were recorded that had been carried out by five Swiss German interviewers following the written German questionnaire. The recordings were recently transcribed, and deviations from the scripted questionnaire were coded according to an elaborate coding scheme. Analyses will focus on variation across interviewers and within interviewers across interviews to identify consistent patterns of adaptation. We will examine the nature and extent of deviations from the script, apparent motivations for such departures, and assessment of whether the adaptations affected the intended meaning of survey items and survey responses.