Use-Files

Scientific Use Files

Researchers working at a university or a publicly funded research institution may receive factually anonymized data sets in form of Scientific Use Files for their research work at their own institute. The term “factually anonymized” means that the data set has been changed in such an extent that a re-identification of the questioned participants by using the provided information is only possible with an excessive amount of time, money and work invested (§3(7) BDSG). Thus, Scientific Use Files de facto do not allow an identification of companies or persons. It is hence not possible to draw any conclusions about an individual company or a single person.

ZEW Scientific Use Files contain neither data on a company’s or a person’s name, nor on addresses and telephone numbers. To support anonymity, data set information that might allow conclusions about an individual company or a single person is aggregated, somehow changed or completely deleted. In terms of the different ZEW Scientific Use Files, data specific anonymization methods are applied.

The Scientific Use Files can not only be used in order to do research at the own institute but also to prepare a potential visit at the ZEW-FDZ. Programming and other preliminary work can already be prepared before the visit.
There are Scientific Use Files for the following data sets:

For data protection purposes, the Scientific Use Files of the Mannheim Innovation Panel and the IAB/ZEW Start-up Panel cannot be accessed by external researchers before the third year following the implementation of the survey.

Education Use Files

Contrary to the Scientific Use Files, the ZEW Education Use Files are absolutely anonymized. It is thus definitely impossible to track back individual information on the participants of a certain survey. ZEW offers Education Use Files of the Mannheim Innovation Panel (MIP) from the survey years 1999, 2001, 2003, 2005, 2009, and 2011. This data is designed for use in teaching at higher education establishments in order to introduce and convey application possibilities of econometric analysis for business data in the context of questions relating to the economics of innovation. The interpretation of the content is, however, limited and thus not suited to answer research questions.

To guarantee full anonymity, different anonymization methods are used radically changing original data. The Education Use File of the MIP represents a resample of the corresponding Scientific Use Files, which has already passed the first step of anonymization. In the resample, the individual data of the metric variables in a data set are draws from the kernel density distribution of the original data and do not correspond with the true statements. The observations cannot be clearly assigned to a specific company. The micro data to be analysed form the basis for generating synthetic statistical units. The artificial data is supposed to carry the same characteristics, i.e. the same multivariate distribution function, as the original micro data. The basic philosophy is to create data sets that are suitable for all kinds of purposes. On the one hand, the passing on of artificial data does not endanger data privacy as no real observations are provided. On the other hand, there is no information loss as long as the given data reflects all distribution parameters of the regarded micro data. (For a detailed description of the procedure, see Gottschalk, S. Unternehmensdaten zwischen Datenschutz and Analysepotenzial, Baden-Baden, 2005. Upon request, a description is also available directly from the author).