Assigning Random Numbers when Co-ordination of Surveys Based on Different Unit Types is Considered (Method)


Sample co-ordination by the use of Permanent Random Numbers (PRNs) is a common method used to have some control over the overlap (number of businesses in common) between samples for two different surveys or between consecutive samples for the same survey. The basic idea is to associate an independent and unique random number, uniformly distributed over the interval (0,1), with every unit in the Business Register. A BR generally consists of several unit types and unit type for a business survey is chosen on the basis of the statistics to be produced. This means that all unit types must be assigned PRNs. There are variuos methods for this but the most straightforward way would be to assign PRNs to each unit type separately. This method means that samples based on different unit types are independent but it does not admit co-ordination between such surveys. The fact that business surveys use different unit types implies a need for this kind of co-ordination. Especially the possibility of negative co-ordination between surveys based on different unit types (in order to spread the response burden) is important when it comes to small businesses.

Another approach to assign PRNs would be to use a method implying that the unit types can be co-ordinated through the PRNs. This method has the advantage to admit sample co-ordination between unit types but, as a drawback, brings dependence between samples based on different unit types. Co-ordination through PRNs cannot meet all objectives of sample co-ordination equally strong and different strategies are discussed in more detail below and references are given to other parts of the handbook.


To read the entire document, please access the pdf file (link under "Related Documents" on the right-hand-side of this page).


Your feedback is appreciated. Please send your remarks, suggestions for improvement, etc. to memobust@cbs.nl.