Titre : | Effects of using an overlapping dual-frame design on estimates of health behaviors: A French general population telephone survey (2017) |
Auteurs : | J. B. RICHARD ; R. ANDLER ; A. GAUTIER ; R. GUIGNARD ; C. LEON ; F. BECK |
Type de document : | Article : Périodique |
Dans : | Journal of Survey Statistics and Methodology (Vol.5, n°2, June 2017) |
Article en page(s) : | 254-274 |
Langues: | Anglais |
Discipline : | EPI (Epidémiologie / Epidemiology) |
Mots-clés : |
Thésaurus géographique FRANCEThésaurus mots-clés BAROMETRE SANTE ; METHODE ; MODELE STATISTIQUE ; POPULATION GENERALE ; ENQUETE ; TELEPHONE ; EPIDEMIOLOGIE DESCRIPTIVE ; PREVALENCE ; COMPARAISONOrganismes INPES |
Résumé : | The Health Barometer (HB) is a repeated general population health survey of the French population. To adequately cover the population, an overlapping dual-frame design of landline and cellular telephone numbers was used in 2014. We assess the overall impact of this methodological change. Data from the HB 2014 were used (n = 15,635). The overlapping approach and the screening approach were compared through several indicators: respondents' profiles, prevalence estimates, response rates, and cost. Dual users reached through the cellular frame were more likely to be males, to be young, to live in urban area, and to have a higher educational achievement, as well as having different health behaviors. Almost all the observed differences disappeared in multivariate analysis. Estimates of health behaviors obtained with the overlapping and the screening approaches were very similar. The highest difference (0.8 percentage point) concerned current smoking. For an identical cost, the overlapping approach is able to reach groups who are underrepresented and unreachable through the screening approach, though producing lower response rates. Health surveys should continue to adapt and explore ways to enhance their data quality. |
Domaine : | Plusieurs produits / Several products |
Affiliation : | Santé publique France |
Accueil