Ein Angebot von
The storage of research data poses challenges for an institution and its technical infrastructure, requiring the creation of long-term and secure storage solutions for various quantitative and qualitative data types. The requirements for storing complex experiments or video files differ from those for storing social science survey data, that is why potentially different storage technologies may be necessary to ensure fast access to different data types.
Within research data management, a distinction must be made between short-term, medium-term, and long-term data storage. Short-term storage occurs during the research process as part of data collection and further data processing.
You can provide guidance to researchers on short-term data storage during the research process. This can help to ensure appropriate handling of research data during the research process.
Medium-term data preservation requires an institutional backup strategy. The following questions should be considered for its development:
[Source: https://ianus-fdz.de/dateispeicherung]
Answering these questions can assist in developing a medium-term data backup strategy in order to protect your institution from potential data loss.
A long-term storage strategy involves archiving research data. Archiving requires not only secure storage but also comprehensive data description with Metadata, which makes the reuse of the research data possible in the first place.
In terms of long-term storage of research data, some file formats are more suitable than others. The decision for or against a certain format depends largely on data compatibility with other data or software to be used, suitability for long-term archiving, and the ability to convert data if necessary.
A rough overview of the suitability of different data formats is shown in the table below:
Sustainability | Machine Readability | Human Readability | Long-Term Readability | Metadata |
---|---|---|---|---|
very good | with widely used open software | yes, without specialized software | standardized standard | fully included |
good | with well-known and documented software | compressed according to standard procedures but technically yes | long-established or widely established | technical details are included |
average | proprietary standard format | with open software (reliable?) convertible to a higher class | relatively new format | some important (e.g., units) details are included |
poor | self-developed reading software | no | just invented | no details |
For institutional storage of research data, in addition to the appropriate storage strategy, a backup must also be established. For further guidance on this, please refer to the article ‘FDM Consulting: Storage Guidance for Researchers.