Survey (non-)participant profiles of GGS France

Abstract

As nonresponse and survey breakoffs are a threat to the representativeness of population surveys, the representativeness of surveys in times of declining and/or low response rates is a crucial factor for social research data. This research uses ERFI2 which is the second French installment of the Generations and Gender Survey (GGS) as part of the Generations and Gender Programme (GGP), a cross-national research infrastructure devoted to data on family demography and life course trajectories. Sampling from the population of adults aged 18–79 living in metropolitan France, this longitudinal survey of three waves attained an overall response rate of 29% (N = ca. 12,500) in wave one with three in four respondents answering via the Web instead of via telephone. As ERFI2 is linked back to its sampling frame, the Fideli database (the French housing and demographic file leveraging administrative tax information on household structure, household incomes, and demographics of household members), it represents a unique and innovative database to study both participants and non-participants of ERFI2. One part of this research is to explore the linkage quality between ERFI2 and Fideli and the linkage consent depending on key demographics. The second part encompasses to derive participants and non-participants profiles based on key demographics reflecting each phase of the survey lifecycle after sampling. The profiling decomposes in which modes ERFI2 sample members were contacted (via post, e-mail, SMS, or phone calls) and in which mode (Web or phone) they eventually completed the survey or not.