Versions Compared


  • This line was added.
  • This line was removed.
  • Formatting was changed.


Record linkage can be performed for different purposes making it a powerful instrument to support decisions in business and government institutions. Record linkage is necessary at least in two basic cases: data collation and list construction (e.g. maintenance of a statistical business register). Data Collation is applied when a project might require data that are not all available from the same source.  List Construction is applied when a project requires a list of all the members of a population to serve as a sampling frame, the contact list for a census, or for collation with data from other sources. Usually there is no single list of the population, but a combination of several lists can be expected to include all or most of it. Usually, there is some overlap between these lists. A de-duplication is necessary to avoid biasing the sample or census so that each member of the population is included only once.

Image Added

Figure 1: Schematic representation of the record linkage process – [4]


Performing record linkage is unthinkable without appropriate software and the development of software packages that implements computational models for record linkage and statistical matching has increased over the last decade. It is possible to find several software packages that perform similar tasks. Such packages may improve the process and its results but, however a lot of time and effort is necessary to decide which one better meets the Organization needs (see for example [7]).