Research Data

The German Research Foundation, DFG, means by research data

[…] digital and electronically storable data generated in the course of a scientific project, e.g. by means of studies of sources, experiments, measurements, surveys, or interviews.

More and more, research funding agencies demand that research data should be archived and accessible for reuse for at least ten years. You can archive and publish your research data on the University of Regensburg Publication Server in a citable manner, either open access or with restricted access.

Research policy increasingly focuses on research data (see Principles for the Handling of Research Data of the Alliance of German Science Organisations). That's why publication of research data will be required by an increasing number of funders and therefore will become more and more like a "normal" publication, showing the scientific output and augmenting the scientific reputation. When published in a citable manner, research data can make your research even more visible. Thus, publishing research data is beneficial not only for the whole scientific community, but it is also in the researcher's own interest.

1. Log in to the publication server and create a new item.
2. Select as item type
• Dataset for a collection of quantitative data (collected, measured, or simulated data or images).
• Experiment for a description of an experiment (e.g., description of the experimental setup, sketches, lab books).
• Image for one ore more image files.
• Video for one ore more video files.
• Audio for one ore more audio files.
• Software for scripts, source code and the like.
• Other for something that is not covered by the other categories.
3. In the "Details" section, enter the metadata for your entry. Mandatory fields are creators (authors), contact Email address, title, and date (at least the year).
• As creators, list all those persons who were directly involved in generating the data.
• At Superordinate Entry, you can link the current (research data) entry with a publication that is based on this data. Enter the item ID of the superordinate entry and press "Lookup". (You can find the item ID of an entry in the table at the landing page, or as part of the URL: http://epub.uni-regensburg.de/12345/.)
4. The "Status" section should contain
• Status: "Unpublished" if you didn't publish your data elsewhere, e.g., in another repository. Even if you have published an article with graphs visualizing your data, the raw data itself is still unpublished.
• Refereed: "No, this document will not be refereed" should be the standard value for research data, since the data itself usually is not refereed during the peer review process of a publication.
5. Select your institutional affiliation (e.g, your working group) and your subject area (up to three).
6. In the "Upload" section you can upload one or more files. Either select a file from your computer, or enter a URL to the file. Please note the section about reasonable file formats! If your dataset contains many files, you may combine them into a ZIP archive.
7. You can enter individual metadata for each file by pressing "Show options" at the respective files. This is particularly helpful if your entry contains several separate files. In these options, you can
• enter a short description that will be shown next to the file,
• determine who is allowed to download the file ("Visible to"). See below for more information.
8. In the "Deposit" section, press "Deposit Item Now". Your entry then is moved to the review workarea of the publication server, will be reviewed by formal criteria by a library employee and then published to the live archive.

For long term preservation, you should choose file formats which are, presumably, still readable in ten years, i.e., no formats that require a specific version of a specific (proprietary) software. For example, you could archive interview transcripts in text or PDF format instead of MS Word files, or measurement data in text or CSV format instead of Origin files.

To ensure citability of the research data entry, we can mint DOIs (Digital Object Identifiers). In the future, research data entries on the publication server should automatically get a DOI. In the meantime, however, please contact us if you'd like to have a DOI for your research data (contact person see below).

DOIs for entries on the publication server always look like 10.5283/epub.12345, where the digits after the dot are the item ID of the respective entry. You can reach the entry always with the URL http://doi.org/10.5283/epub.12345, even if the URL of the publication server changes.

The University Library Regensburg offers a self-learning course on research data management (in English) on the e-learning platform G.R.I.P.S.:

• Information on the online course Research Data Management Training

You can find more information on the Research Data page of the University library of Regensburg (in German).