Statistical matching: a model based approach for data integration - Products Statistical working papers

This publication provides a general introduction to statistical matching with an emphasis on implementation in an applied context, namely within the system of European social surveys. Statistical matching is a relatively young statistical technique that aims at integrating information from different data sources. As such it can enhance the value of existing data sets or even contribute to a reduced burden for respondents. The publication presents in detail the methodology and results from two pilot studies focused on specific needs within the framework of social statistics: (1) the provision of joint information on various quality of life indicators; and (2) labour statistics enhanced with wage information. Based on these studies we draw conclusions on the quality of the results and the potential usability of this technique.