Language questions and statistical surveys


Vera Prosdocimo, Julia Valle, Yohan Bühler

09.2014 - 08.2015
Diversity, Evaluation, Institutions, Interaction, Policy

In collaboration with the Federal Statistical Office (FSO)

In 2010, the Federal Statistical Office (FSO) began conducting an annual population census. The data are mainly gathered from registers of residents, with additional information coming from survey sampling. Five politically relevant issues are examined in depth in yearly thematic surveys; in 2014, FSO conducted its initial survey on language, religion and culture, which will be repeated every five years in future. In our work, we analysed the survey section “language” to uncover the various challenges inherent in a quantitative linguistic survey in multilingual Switzerland. The project focused on the following questions:

  • According to which criteria and with which objectives were the language-related questions drafted in the 2014 survey?
  • What challenges arise when formulating questions and explanations?
  • What types of negotiation processes arise during telephone interviews?
  • What questions and terms cause the interviewees comprehension and interpretation problems?
  • What regional language differences can be observed?

To answer these questions, our work comprised ethnographic research and document analysis, including an investigation of the documents that FSO and the mandated social research institution compiled before and after the 2014 survey (questionnaires, schooling material and follow-up instructions for the interviewers in German, French and Italian). In this context, we also attended training sessions and de-briefing events held for the interviewers. To accommodate our research demands, the mandated social research institution and FSO conducted a parallel survey of roughly 150 individuals (50 interviewees in German, French and Italian). These telephone interviews were recorded with the express consent of the interview partners. Over the course of this parallel survey, we were able to accompany and meet with the interviewers at three sites, each located in a different linguistic region.

Purpose – Expected results

Our analysis of the questionnaire’s conception and its usage in telephone interviews revealed both opportunities and limitations of an in-depth survey in the context of increasingly multilingual practices. Despite meticulous preparation of questions, terms, explanations and filters used as well as thorough training of interviewers, there still remained some room for interpretation and manoeuvre for everyone involved. Analysis of telephone interactions and focus group interviews showed that both interviewers’ and interviewees’ beliefs and knowledge on language highly influence the gathered data.