Abstract
In 2012, a new national survey among homeless populations was carried out in France by INSEE and INED, 11 years after the first one. According to an indirect sampling design, users of services dedicated to homeless people (shelters, soup kitchens…) were randomly selected for a face-to-face interview or a short self-administered questionnaire (SAQ). For the first time, non-French speakers were also surveyed in order to improve the field coverage of the survey and to obtain information about these little-known populations. Thus, the SAQ has been translated into 14 languages supposed to be the most used by non-French speaking users of services, and tested. Eventually, around 5800 SAQ were collected, filled in by natives of 120 different countries.
This presentation will cover the various steps of the protocol focussing on two main difficulties: identifying the “right” languages used by the non-French speaking homeless population; adapting and translating the SAQ, taking into account literacy skills and cultural conventions. Then, a mapping of the languages used by respondents and an assessment of the data quality based on a behaviour coding of the responses will be discussed. To what extent does data quality reflect the literacy skills of respondents or their subjectivity?
confirm funding
Event ID
17
Paper presenter
54 351
Type of Submissions
Regular session only
Language of Presentation
English
Weight in Programme
1 000
Status in Programme
1
Submitted by géraldine.vivier on