Five advice for strategic management of research data
Make a habit of structuring and documenting data
Data is valuable, and especially in major research projects, it is important to have a consensus on how (and where) data is saved. Cambridge University provides advice on how to easily work with folder structure and naming principles to structure collected data. It is also important to document data, for example, who collected it for the future. A starting point can be the archive form ["Documentation of Research Materials"].
Save these information so you can access them later. It is also important to systematically document changes when raw data is processed.
Within some areas of science there are worked up processes for structuring and documenting data, for example in the form of labbooks. Ask your colleagues about how to do in your area!
Take a habit of saving safely
Be sure to back up your data, and sometimes check it if it works.
Consider if you are helped by writing a data management plan
If you engage in extensive data collection, if financiers and ethics committees require it, if you are to collaborate with researchers from other universities, companies or organizations, it may be useful to clarify who owns the data, how it can be used and managed before starting data collection. Some financiers also demand data management plans.
A data management plan is an account of how research data should be generated, collected, processed and maintained. Sometimes it also describes methods and standards associated with data management. SND (Svensk Nationell Datatjänst) has general advice and instructions on how to write a data management plan. Some examples of how to write a data management plan are described on the University of California San Diego website. A good template is DMP Online.
Storage of research data in DiVA
Several journals and research financiers require research data to be made available online. A small amount of research data can be stored in DiVA via the Dataset publication type. Each file may be up to 16GB and you can upload multiple files, even in compressed formats. Each dataset receives a unique identifier and a persistent link. From the dataset it is possible to link to the publication and on the contrary to indicate which publication data sets belong to. In connection with this you should also attach a documentation containing a description of research data e.g. an overview of the research results, various variables that are included in the data material and what they mean, how the material is to be interpreted, etc. For more information send a mail to firstname.lastname@example.org.
LiU currently has no official policy regarding handling and storage of research data.
Citation of research data
Data sets can be cited provided that data sets have a persistent identifier such as DOI or urn:nbn. As a member of CrossRef, we can help to assign DOI to data sets to make it cited.
For other questions about research data, we refer primarily to the archive.