This post is also available in:
Deutsch
For almost two years now, the SNSF has been requiring researchers to submit a data management plan (DMP) with their funding application. We asked the SNSF: How do you rate the quality of the DMPs submitted so far? The positive answer: 70% of the DMPs are accepted without further comments. The most frequent reasons for revising the remaining 30% are: Insufficient explanation of how data is shared, unjustified embargo periods and lack of definition of metadata. In addition, it would make the work of reviewers easier if a clear structure of DMPs was ensured.
In the following, we summarize 5 tips on how to avoid the most common errors.
- Describe precisely why data cannot be shared. It is legitimate not to share certain data, for example, if it is sensitive or personal non-anonymizable data, or if data is subject to copyright restrictions. However, even in such cases, it may be possible to share the metadata about such data.
- If necessary, define a reasonable embargo period. Embargo periods allow scientists in certain cases to withhold data for a certain period of time. Typically, embargo periods of 6-12 months are acceptable.
- Use existing metadata standards. It is essential to use existing metadata standards to ensure that data are findable. A list of technical metadata standards is available at http://rd-alliance.github.io/metadata-directory/. If a standard has not yet been described, orient yourself to similar studies and define a standard for your project.
- Keep it simple. The DMP should be clearly structured. For example, it has proved useful to create a table of all data and number the data types consecutively so that you can refer to them in the following chapters.
- Call for help. In addition to helpful documents from the SNF and the accompanying questions on the DMP form under mySNF, researchers at the UZH can also make use of individual support. The Data Services team at the main library regularly offers workshops on writing DMPs and conducts reviews of DMPs. If you have any general questions on how to handle your data, write DMPs or if you want us to review a DMP, please contact us at data@hbz.uzh.ch.