Problems Associated With Designing Subannual Business Surveys

Methodological problems associated with designing continuous subannual business surveys tend to be more complex than those encountered in other surveys. The complexity is primarily due to the highly dynamic nature of the sampling frame. The purpose of this article is to highlight these problems, especially those relating to the creation of the sampling frame and to sample rotation, selection of births, and elimination of deaths. Three methods of sample rotation are described. Some practical solutions to these problems are proposed to ensure that the sample reflects the structure of the population on any occasion.