The DACSEIS Project

Abstract

Data quality has become increasingly important in social and economic statistics, especially when data are gained from various sources obtained using different methods. In many cases, measuring data quality is strongly connected to variance estimation as one error source. The DACSEIS project investigates variance estimation methods for complex surveys in the context of an applicable and comparable quality measurement of European household surveys.

The aim of the project is to deliver recommendations of how to measure data quality in complex surveys that take various aspects of error sources into consideration and allow the handling of different components of the measures. This will include a complex Monte Carlo simulation study in a practical environment, e.g. in synthetic, but realistic, universes that are close to selected European household surveys, to test the wide range of theoretical methodology in practice.