Abstract

This presentation will give an overview of skills and methods needed to implement new data sources in the domain of Official Statistics using a few examples from Statistics Netherlands. Not only does the introduction of some of these new data sources require new skills (and/or new staff), but there's also impact on more classic methodology (e.g. combining data sources, non probability sampling, concept drift). And since these new data sources are more loosely coupled with the (often prescribed) output, the combination of methodological and content knowledge, as well as the flexibility of statistical production process have grown in importance.