Not all research data generated in the course of a project is worth archiving or publishing. Especially in the case of large amounts of data, a high storage requirement for archiving/publication can also lead to a higher administrative effort and consequently to higher costs. It is therefore advisable to decide early on what should be done with which data.
The 5-step guide of the Digital Curation Center (DCC), for example, can help with the decision.
1. Identify the purposes that the research data could serve.
2. Identify research data that must be retained (e.g., through contractual requirements).
3. Identify research data that should be retained (e.g., data that could have long-term value).
4. Weigh the costs.
5. Complete the data assessment.
Always consider which data must be kept and for how long, or which data must be deleted and why (e.g. personal data).
An example of a data set that should be preserved and made available to the public is research data that is unique or very difficult to reproduce and that may be of good use to the research community. Research data produced for testing purposes only, on the other hand, do not need to be archived or published.