Methodenreport: Synthetische Scientific-Use-Files der Welle 2007 des IAB-Betriebspanels

Abstract

"Providing scientific use files for business surveys is a difficult task. Due to smaller populations, higher sampling rates, and skewed distributions disclosure risks are much higher than for household surveys. Simple measures like coarsening are not sufficient to protect the data. The aim of generating synthetic datasets is to release data that provide a high level of data utility while guaranteeing the confidentiality of the survey respondent. To achieve this, sensitive variables and variables that could be used for re-identification purposes are replaced with multiple imputations. This report gives a short introduction to the topic and discusses some aspects that analysts should keep in mind when using the synthetic datasets. Furthermore, the report describes how valid inferences can be obtained based on the synthetic datasets and provides some first data utility evaluations that indicate the potentials but also the limits of the generated datasets." (Author's abstract, IAB-Doku) ((en))

Cite article

Drechsler, J. (2011): Methodenreport: Synthetische Scientific-Use-Files der Welle 2007 des IAB-Betriebspanels. (FDZ-Methodenreport 01/2011 (de)), Nürnberg, 19 p.