Quality Framework for the European Working Conditions Survey

The European Working Conditions Survey (EWCS) has been a cornerstone of labour research for over three decades, but with over 56,000 interviews conducted across 29 countries, ensuring consistency and data quality is a serious challenge.

For the 2024 wave, rowsquared independently assessed the quality of all collected interviews and designed a data-driven quality framework to determine which interviews should be included in the final analysis. We classified interviews based on clearly defined quality indicators, and developed a data quality framework to guide future waves.

rowsquared developed a R-based workflow to assess face-to-face interviews. We identified low-quality interviews by looking for patterns such as straightlining or response similarity in process data like timestamps, question duration or routing. Each interview was assigned a composite quality classification to guide inclusion or exclusion.

Our data quality framework provided a replicable model, including a Normalized Anomaly Index, that Eurofound can use in future waves. We reviewed data from Eurofound’s data provider and flagged quality issues. We provided recommendations for integrating live data quality monitoring in future collection cycles, and advised on how terms of reference for data providers could better reflect evolving quality assurance expectations.

By identifying and removing low-quality interviews from the 2024 dataset, we improved the reliability of the EWCS data. Our work laid the foundation for data-driven quality management, enabling Eurofound to detect quality issues earlier, and deliver more robust survey outcomes.