| Presenter(s): | Eva Endres & Thomas Augustin, Ludwig-Maximilians-Universität München |
| e-mail(s): | eva.endres@stat.uni-muenchen.de augustin@stat.uni-muenchen.de |
| Date: | 4 April 2018 |
| EMOS learning outcome addressed | Statistical methods |
| Webinar aims | The combination of information from different available surveys is a task which can be solved by statistical matching. In contrast to record linkage, statistical matching integrates the information of data files which are based on different observations. This webinar gives an overview on the statistical matching problem and how it can be solved. A special focus is set on conditional independence which forms the basis for a group of statistical matching procedures. |
| Webinar learning outcomes | To understand fundamental aspects of statistical matching, including its aim and limitations. To be aware of the crucial role of the conditional independence assumption in matching. To gain insights into selected micro and macro approaches for statistical matching. |
Webinar content | – Basics, framework and aims of statistical matching. – Conditional independence. – Selected micro and macro approaches for statistical matching. – Framing the statistical matching problem as a problem of missing data and imputation. – The maximum likelihood approach in statistical matching. |
| Difficulty level | Introductory |
| Prerequisites for the webinar | Basic knowledge of probability theory and statistics is sufficient. |
| Further readings and resources | D’Orazio, M., Di Zio, M. and Scanu, M. (2006). Statistical Matching – Theory and Practice, Wiley, Chichester, United Kingdom. |
| Presentation material | EMOSWebinar07_handout |
| Recording of the webinar | EMOSWebinar07-Recording |
