enEnglish
CROS

WP2: Development of methods

Coordinator: CBS

Start date:  March 2010
End date: July 2011
WP status:  completed

Here you can find the final report

WP description
This WP deals with those methodological developments that should be obtained in order to make record linkage, statistical matching and micro integration processing applicable in the ESS. The key statement behind the proposed sub tasks is the use the proposed methods in a national statistical institute, that needs to:

  • assess result quality,
  • reuse data integration results for any statistical purposes (imputation, editing, disclosure control, weighting, estimation)
  • take complex survey designs into account, if some of the sources to integrate are sample surveys, modifying the methods proposed for samples consisting of independent and identically distributed observations.

The proposed methods are complemented by examples that can show how they can be applied in real life cases.

The report is organizaed in the following chapters:

  1. Application of Bayesian record linkage in a real life context: a comparison with traditional record linkage approaches and comments on the analysis of linked data
  2. Editing errors in the relations between units when linking economic data sets to a population frame
  3. Methodological developments on statistical matching
  4. Handling incompleteness after linkage to a population frame: incoherence in unit types, variables and periods
  5. Bootstrapping Combined Estimators based on Register and Survey Data
  6. Models and algorithms for micro-integration
  7. Macro-integration techniques with applications to census tables and labor market statistics