Preserving and publishing data

Tip: Why is this topic important
» Your data is valuable – they should be available for reuse.
» Archiving data is also a way to ensure availability of data for yourself at a later time point, e.g. after switching affiliations.
» Rich metadata will make your data findable through search engines. Good documentation and structured metadata will enable others to understand the information held in your data and enable data reuse.
» Comply with requirements by research funders and academic publishers to making data underlying a study available.
» Increasingly, archiving your data as FAIR as possible is viewed as equally important as publishing other research output, in the assessment of research.

About this chapter

Research data should be archived and preserved, for the benefit of future research. Data should be archived as open as possible and as closed as necessary, as the guiding principle of EU and others says. Not all data can be made publicly available, as elaborated in the chapter “Legal and ethical aspects”. In this chapter you will be asked to consider whether you data, or part of your data, may and should be preserved and published, and to explain why access to (part of) your data need to be restricted. The chapter will assist you in walking through all planning considerations regarding these issues, including where to publish your data and what metadata to apply.

Question-specific guidance

Selecting archive(s) for publishing datasets

add figure for flow chart

This question, and its sub-questions, will assist you in finding archive(s)/repository(-ies) suitable to deposit your data.

Where do you plan to archive you research data? Remember that this can also include project results that you are not immediately thinking of as ‘data’. As archive choice often decides over metadata standards to be followed, it is advisable to investigate this question early in the project.

Timely archiving of datasets in trustworthy repositories is recommended. Depending on the study and the type of data, it may be appropriate to archive everything in one repository or archiving different parts in different repositories that will link to the related datasets. Re3data and FAIRsharing are registries that may help identifying suitable repositories.

Evaluate which archive(s) will be most relevant for your datasets, using the following the decision tree:

Does your dataset contain personal data or sensitive information?
[If yes, investigate discipline-specific or generic archives with restricted access]

Do journals or funders require that specific archives will be used or is there disciplinary conventions?
[If yes, make yourself familiar with the recommendations and use discipline-specific archives]

Can you use an institutional archive?
[If yes, use your institutional archive]

If none of the above applies, use a generic research data archive.

In some cases it is important to contact the research data archive beforehand. elaborate

Can all of your data become completely open over time?

The issue here is for you to consider whether (some of) your data is of a nature that should not be made openly available. Please see the chapter ‘Legal and etical aspects’ to identify criteria for deciding this. This should help you identify whether part of your data need to be restricted from public access. Your will also be asked to consider whether such parts may be anonymized or aggregated so that such a version of your data can be openly available. Data that cannot be openly available, may perhaps be available to authorized users. You will be asked to consider this option, and how such authorization may be carried out in practice.

Securing that your data is FAIR

No matter whether your data may be published openly or not, it is always a question of making your data FAIR - or rather as FAIR as possible. FAIR stands for Findable, Accessible, Interoperable, Reusable. So you should aim to make your data as findable, as accessible, as interoperable, and as reusable as possible. So here you will find a checklist that serves to remind you how to make your data as FAIR as possible.

List data(sets) that you will delete

Determining whether data need to or should be deleted, is a decision you may need to do. There may be data that were ruined for some reason. Or other data, in huge amounts, that obviously are of no value for anyone. You need to decide whether there are data that are of no value to preserve. In a sub-question here, you will be asked to explain why these data were deleted. This explanation is part of making your research transparent.

List datasets (and software or code) that you will preserve or publish

Here you will be asked to describe closer what data you will preserve. The question of sensitive information will be asked, to ensure that all necessary considerations regarding this issue is done. You will also be sked about the metadata and documentation to follow the data. Metadata and documentation are very important for the preserved data to be of good value into the future. So you are asked to think through and consider this well.