Opening research data

Äänekoskella tarjottaviin syksyn avoimiin korkeakouluopintoihin ilmoittaudutaan nyt. Tervetuloa mukaan!

Opening research data

The Ģ��ֱ�� strives for the widest possible transparency of research data and their descriptive information. On this website, you will first find general information on the openness of research data and then concrete instructions on how to open them.

Principles

The Ģ��ֱ�� follows the internationally accepted principle of "as open as possible, as closed as necessary". For each dataset, the aim is to be as open as possible, but as restricted as necessary - simply wanting to keep the data closed is not sufficient justification.

According to the Ģ��ֱ��'s data policy,

According to the Ģ��ֱ��'s research data policy, at least the descriptive metadata are always opened.

Opening must always be in line with FAIR principles: the researcher hands over the data to an appropriate, trusted body such as a university or archive, which labels it with descriptive information and gives it a permanent identifier (most prominently, DOI). The body thereafter ensures that the data is reliably accessible by means of the identification link and the permanent landing page. The service is also responsible for the distribution and long-term preservation of the data.

Levels of openness

'Opening' means making the data available either at the level of descriptive data, in an openly downloadable format or in a licensed form with limited access rights. All of these can be accompanied by various embargoes to ensure that the researcher has the right to access the data first. There are several degrees of openness, depending on the nature and content of the data and the needs of the researcher.

If the data is not suitable for publication and no suitable archive is currently available or archiving is not possible for any other reason, it is always possible to publish the basic description data, or metadata, of the dataset. This is also often the minimum requirement for the transparency of data by funders and research institutes.

In the case of metadata, publication as early as possible is recommended - even if the research is still ongoing. A metadata publication tells the world about the researcher's expertise and the existence of the research data even before publication! One of the main functions of metadata publication is to increase the visibility of the researcher and the benefits are greatest the longer the metadata are available while the research is still ongoing.

At the Ģ��ֱ��, descriptive data is published using the Converis data section, and the JYX repository data catalogue serves as the publishing platform. Metadata meet the FAIR criteria when published by an established, standardised machine-readable descriptive language descriptive catalogue such as JYX or Etsin. The publication service provides the researcher with the tools to compile the descriptor. They, in turn, are experts on their data and know how to describe it in a way that matches the data and is informative.

Remember that metadata can be used to publish documentation describing the data and the research, such as the structure of the questionnaire and interview questions, even if the datasets themselves cannot be published due to legal or ethical reasons.

Anonymous and otherwise non-sensitive data, to which the researcher has rights, can be published openly online. In this case, the researcher chooses a trusted, industry-standard publishing service that provides at least a landing page and a persistent identifier (e.g. DOI or URN) for the data. The Ģ��ֱ��'s JYX repository is a good suitable publication service. Publication means that the publication service meets the FAIR criteria. For example, publication on a project or personal website does not therefore constitute publication. Openly published data is always subject to a standard licence that governs its use.

If the data contains personal data or is otherwise sensitive, it is usually possible to either open it in the repository only with a licence and limited access to archive it, i.e. place it in a repository that is responsible for its storage, curation and distribution. Access to the dataset may be subject to various conditions. In most cases, this means restricting the distribution of the data to research use only - possibly only for research on specific topics. In this case, the publisher of the material publishes the basic metadata of the data and interested parties contact the publisher of the data, who takes care of licensing and distribution.

Note that self-storage by the researcher in university storage systems or, for example, on the hard drive of a personal device, is not archiving, but ongoing storage.

What can, and should, I publish?

There are many types of research data, some of which are suitable for publication as such, some of which cannot be published at all and some of which can be adapted for publication. In principle, data containing personal data or confidential information should not be made openly available. However, in the case of data that can be used for other research purposes, it is possible to publish the data behind restricted access. For example, data containing nesting data on endangered animals may not be made publicly available, but may be used for research purposes. Similarly, third-party rights (e.g. copyright) to the data may prevent publication unless otherwise agreed.

It is also worth thinking as broadly as possible about your research data, so that you may find data to publish from unexpected directions. This may include codes, corpora, literature reviews, derived data, etc. In addition, even when the results themselves cannot be published openly, related documentation can be published to provide further insight into the data and related research (, ).

For example: In historical research, original data and materials are governed by archives, which are responsible for their preservation and distribution, and may also manage the rights to them. They cannot be republished. In contrast, databases, corpora and other derived data compiled from these materials can be managed and potentially published by the researcher.

Data can also be published as different entities, which may be in hierarchical relationships with each other in publishing systems. It is possible to describe large entities at the level of metadata alone and to publish different functional entities under them as sub-entities. For how to proceed, see the 'Creation of metadata in Converis' instructions site listed on the bottom of this page.

What can I open, and what should I open? Organise your data in advance so that you know which parts are suitable to open and which to dispose of as planned. If you are using third-party data in your research, such as data from another researcher, an archive or a registry, your initial access agreement should outline what you will do with the data once it has been analysed. As a general rule, the terms of use require you to destroy copies of the data when your use them for your research. If the data you receive from a third party contains direct or indirect personal data (including pseudonymised data), destroy the copies of the data.
Anonymous, non-personally identifiable data that does not contain personal data and does not involve legal or ethical factors that restrict its disclosure, such as trade secrets, may be made openly available in a suitable repository whenever convenient. If your data contains personal data, follow the lifecycle plan indicated in the privacy notice to either anonymise the data, destroy it, or make it available in identifiable, restricted form in an appropriate archive. Anonymous and anonymised data can be published in JYX or any other trusted, FAIR-compliant repository known in your field.
By the time you intend to archive or publish data, ensure that they are accompanied by all the documentation necessary for its proper interpretation and purposeful use. Without documentation and an explanation of the content and, where appropriate, the structure of the data, even high quality data will become unusable in no time!
Where should the data be opened? The Ģ��ֱ��'s publication repository JYX is an excellent solution for opening anonymous, non-archival data. If there is a repository or archive commonly used in your field (e.g., the Finnish Language Bank, CESSDA or the Finnish Social Science Data Archive, Dryad, Zenodo for code publishing), you should choose it.
When the data is to be published, choose an appropriate licence, e.g. Creative Commons 4.0 Attribution-International. The Open Science Centre or your chosen repository will help you to choose the licence.
When do I open my data? Anonymous, non-archival datasets can be published as soon as they are ready, i.e. when there are no further changes to the files and their descriptive documentation. If the data contains personal data, the life-cycle plan indicated in the privacy notice is to be followed.