The French Generations and Gender Pilot Survey: Internet versus Telephone, or Both?

Ruxandra Breda-Popa, INED
Milan Bouchet-Valat, French Institute for Demographic Studies (INED)
Arianna Caporali , Institut National d'Études Démographiques (INED)
Géraldine Charrance, INED
Efi Markou, INED
Thomas Merly-Alpa, INED
Laurent Toulemon , Institut National d'Études Démographiques (INED)

The Generations and Gender Programme (GGP) questionnaire has a long and complex structure. Its adaptation to the web poses challenges in terms of low response rates, self-selection or inaccurate responses. The French GGP2020 test compares CAWI (computer-assisted web-interviewing) and CATI (computer-assisted telephone interviewing) modes, the latter having often been implemented at INED. The test includes nine scenarios, making it possible to measure the impact of financial incentives (pre- and post-interview) on the response rate and the continuation of the questionnaire or the interview. One of the scenarios offers respondents CAWI mode alone. The other eight scenarios combine CAWI or CATI, presence or absence of unconditional and conditional incentives, the latter offered when the questionnaire is completed. We also test the efficiency of offering a new response mode (switching from CAWI to CATI, and from CATI to CAWI). After the test, we will decide our data collection mode, likely a mix of CAWI and CATI in order to control for mode effects. We use a Demographic file on housing and people based on tax data (called Fidéli) as a sampling frame. Beyond GGP, it will be interesting to highlight lessons that can be drawn for other French surveys, for example concerning the possibilities of using Fidéli’s e-mail addresses for a “push-to-web” survey on a representative sample, and the impact of financial incentives in France. For GGP, this test will also focus on the harmonized questionnaire of the second data collection cycle, which has been translated into French by INED.

See extended abstract

 Presented in Session P1. Postercafe