Publication


Arnaud Régnier-Loilier
Inconsistencies in the Number of Children Reported in Successive Waves of the French Generations and Gender Survey
Population, 2014
URL, JabRef BibTex, Abstract
What is the best way to ask survey respondents about the number of children they have? Arnaud Régnier-Loilier compares the number of children reported by the people who completed the three waves of the Étude des relations familiales et intergénérationnelles (ERFI), the French version of the European Generations and Gender Survey. While there was no single question in the first two waves about the number of children born or adopted, this figure was determined by a succession of questions about the number of children alive or deceased, resident or non-resident in the household. This procedure introduced errors which are measured in this article by a further question contained in the third wave which asked respondents to give their total number of children. It shows that some respondents tend not to report non-resident children. The article draws survey designers’ and users’ attention to these collection biases and highlights the value of including a question about the total number of children.

Reference


@article{Régnier-Loilier2014a,
  author = {Arnaud Régnier-Loilier},
  title = {Inconsistencies in the Number of Children Reported in Successive Waves of the French Generations and Gender Survey},
  year = {2014},
  journal = {Population},
  volume = {69},
  number = {2},
  pages = {167-190},
  url = {http://www.ined.fr/fichier/t_publication/1691/publi_pdf1_population_fr_2014_2_enquete_qualite_des_donnees_nombre_d_enfants_declares_collecte.pdf},
  timestamp = {25.09.2014},
  abstract = {What is the best way to ask survey respondents about the number of children they have? Arnaud Régnier-Loilier compares the number of children reported by the people who completed the three waves of the Étude des relations familiales et intergénérationnelles (ERFI), the French version of the European Generations and Gender Survey. While there was no single question in the first two waves about the number of children born or adopted, this figure was determined by a succession of questions about the number of children alive or deceased, resident or non-resident in the household. This procedure introduced errors which are measured in this article by a further question contained in the third wave which asked respondents to give their total number of children. It shows that some respondents tend not to report non-resident children. The article draws survey designers’ and users’ attention to these collection biases and highlights the value of including a question about the total number of children.}
}

Subscribe to our Mailing List!

Fill the form below with your contact information to receive our bi-monthly GGP at a glance newsletter.