Selecting kindergarten children by three stage indirect sampling

Abstract

"One cohort of the German National Educational Panel Study will consist of a sample of kindergarten children. No nationwide frame of kindergartens is available in Germany, contrary to the situation for primary schools. Following the works of Lavallée, we present a solution by indirect sampling, using links between kindergartens and primary schools. In the first stage, primary schools are randomly selected and all kindergartens are identified that have a link to the sampled schools. In the second stage, for every sampled school we take a random sample of kindergartens linked to this school. In the third stage, samples of children within the sampled kindergartens are selected. Using the links of the sampled kindergartens to all schools in the population, unbiased estimation for the population of children in kindergartens is possible. Our second stage sampling is in addition to the existing literature." (Author's abstract, IAB-Doku) ((en))

Cite article

Kiesl, H. (2010): Selecting kindergarten children by three stage indirect sampling. In: American Statistical Association (Hrsg.) (2010): Proceedings of the Survey Research Methods Section, American Statistical Association (2010) (Proceedings of the Survey Research Methods Section, American Statistical Association, 2010), p. 2730-2738.