2  Data use in the FDZ infrastructure

2.1 Workflow from the analysis to the release of the results

As a rule, the data should be used in Stata. If individual analyses are not feasible with Stata, it is possible to work with R or Octave. To do so, please contact the FDZ in advance.

The FDZ’s data are analysed through controlled remote execution, during an on-site use or remote desktop access. For both access methods, the FDZ uses the browser-based software JoSuA (Job Submission Application). For more information on JoSuA and the different modes, please refer to the manuals. Please use the Internal Use Mode in JoSuA for data preparation1, preparation of on-site visits and for testing the runnability of the evaluation programmes. The results visible to you in the Internal Use Mode are censored through an automatic data protection check. You can look at the results on the JoSuA web interface, but you are prohibited from downloading them 2. The data sets stored in your project directory in the Internal Use Mode are also available in the Presentation/Publication Mode. Please note that the datasets saved in remote desktop access must be generated again via JoSuA.

Please upload the programmes of the selected results that you want to publish in a presentation or publication in the Presentation/Publication Mode. Before doing so, please make sure that your do-files run correctly in the Internal Use Mode and that the number of observations underlying the results is sufficient 3(further information on the required number of observations can be found later in this document). The programmes and result files uploaded in the Presentation/Publication Mode are subject to a data protection check by the FDZ team.

As a rule, no further manual deletions or censoring are performed during this process. In limited exceptional cases, the FDZ carries out manual deletions in the log files. For this purpose, sufficient reasons why the results cannot be displayed in a different form must be provided in the programme files. After successful review by the FDZ, the programmes and result files will be made available for download. If critical results are still contained, the affected log file or, if applicable, the entire affected job will not be released. You then have the option of revising the corresponding programme file accordingly and uploading it to JoSuA again.

2.2 Contractual obligations and sanctions

Despite the checking routines in the FDZ, you are obliged to check the transmitted results again yourself. In case of doubts, the results must not be published, but reported to the FDZ. In addition, you agree to maintain confidentiality in the data use agreement, which includes, among other things, that no re-calculations will be made to determine deleted values. In case of violations, you or, in severe cases, your entire institute or university will be excluded from any further data use for a period of up to two years. Furthermore, the penalties specified in the use agreement may apply. Information about violations of the above obligations will also be passed on to other research data centres.

  1. At the end of the project, you can upload the data preparation programmes via the Presentation/Publication Mode for verification requested by reviewers of journals. Please note this in the JoSuA comment field.↩︎

  2. The results generated in the Internal Use Mode may not be copied, photographed, printed or duplicated. It is also prohibited to allow third parties to see them. Any use of these results except for the further development of your programmes constitutes a breach of contract.↩︎

  3. We reserve the right not to release any result files within JoSuA jobs that have obviously not first been tested in the Internal Use Mode for runnability and compliance with data protection.↩︎