Frequently asked questions about open science

Open science is an approach to scientific knowledge production that covers the entire research process. This includes, among other things, open applications, open methods, open-source code, citizen science, open peer review, and open evaluation. On the university’s website, the focus is on open publishing and data sharing, as these are the areas where the library and the research support services can most effectively assist academic staff.

Preprints are complete drafts of scientific articles intended for publication in peer-reviewed journals. Preprints are not peer-reviewed. They are primarily published in dedicated repositories for this purpose, well-known examples being arXiv and bioRxiv. Publishing preprints is most common within the natural sciences. Different publishers and journals have different policies regarding preprints. Most open access journals accept manuscripts that have previously existed as preprints, but they typically require authors to acknowledge the existence of the preprint to clearly distinguish it from the final, peer-reviewed article. Some journals also require that an article must be substantially different, in terms of adding new knowledge, from the preprint to be considered and published. Preprints are often assigned DOIs, which also indicates that they are not deleted even after the articles based on them have been formally published.

Preprints are controversial. Advocates of preprints often point out that this practice enables so-called “community feedback.” In this way, article manuscripts can be further improved before they are submitted to journals for review. Some also emphasize that preprints reduce the risk of conflicts in cases where questions of priority arise, and that they can generate attention around research results more or less immediately, rather than waiting several months for the journal publication process. This may increase citations for related peer-reviewed publications in the future.

One of the main concerns about preprints is that some researchers may “settle” for publishing preprints and never submit them for review and publication in journals. Over time, this could weaken the criteria for what is considered rigorous research. Others are concerned about what this practice may mean for qualitative research based on personal data, as it may be possible, and highly problematic, to publish research that has not undergone the ethical scrutiny of peer review.

Sharing research data and scientific publications helps make research more accessible, verifiable, and useful for others. When results and data are open, they can be checked, reused, and built upon by other researchers.

Open access can also increase the visibility and impact of research, as more people can access it without financial barriers. Data sharing can also save time and resources by reducing the need for new data collection and improving quality by enabling comparison and verification of results.

At the same time, sharing facilitates collaboration across disciplines and institutions and makes it easier to develop new knowledge based on existing research.

For the individual researcher, open sharing may also have concrete benefits. When publications and datasets are freely available, the likelihood increases that the work will be read, used, and cited by others, which in turn can lead to new collaborations and research opportunities. In addition, open publishing and sharing of research data are increasingly emphasized in the evaluation of research and research careers, both nationally and internationally.

There are several models for open publishing, which differ in how access is provided and who pays for publication.

• Gold open access means that the publication is made openly available in a journal, often against payment of a publication fee
• Green open access means that a version of the article is made available in a repository, for example through self-archiving in NVA
• Hybrid open access refers to subscription-based journals where individual articles can be made open upon payment, while the rest of the content remains behind a paywall
• Diamond open access involves publishing in open journals that charge neither readers nor authors

The different models have different advantages and limitations, and which is most relevant depends on factors such as the publication channel, funding, and institutional agreements.

You can use the Journal Checker Tool to investigate whether a journal meets requirements for open publishing from funders such as Plan S. The service allows you to search for journals and see what options exist for publishing open access, either through publishing agreements, open journals, or self-archiving.

For research to be shared and reused, both articles and datasets must be assigned a license that specifies what others are allowed to do with the material. A license regulates, among other things, whether it can be shared, reused, or adapted.

For scientific articles, standard licenses related to publishing are often used, particularly Creative Commons licenses. For research data, it is recommended to use open licenses, and at Østfold University of Applied Sciences, Creative Commons is recommended. A particularly open license is CC0, which effectively makes data freely available for use.

Even if a license such as CC0 allows free use without requiring attribution, fundamental norms of academic practice still apply. The Vancouver guidelines, which are standards for citation and referencing in scientific work, mean that sources must be credited even when the material can in principle be used freely.

The choice of license has significant implications for how material can be used. Restrictive licenses may limit reuse and make it difficult to combine datasets from different sources, while open licenses better facilitate further research.

There are several ways to investigate this. A good starting point is to look up the journal in the Norwegian Register for Scientific Journals, Series and Publishers.

If you know which publisher produces the journal, you can also check the overview of national publishing agreements. Select the publisher in the list and open the list of participating institutions for the current year. The title list provided there shows which journals are included in the publishing agreement.

Subscription journals that offer open publishing upon payment are often referred to as hybrid journals. In these journals, most articles are published behind a paywall, but individual articles can be made open by paying a publication fee.

The university’s publication fund does not support such publishing for two main reasons. First, it involves a form of double payment, where institutions both pay subscription fees for the journal and publication fees to make individual articles open.

Second, hybrid journals are not based on a principle of full open access. Open publishing as a research policy goal is about making entire publication channels open, not just individual articles. Therefore, support is prioritized for publishing in fully open journals.

It is possible to share research data containing personal data, but it requires thorough legal and ethical assessments. Data protection legislation sets requirements for, among other things, data minimization and risk-reducing measures, and data should be anonymized where possible. If anonymization is not possible, data should be pseudonymized. Such data are still considered personal data, as it may in principle be possible to re-identify individuals.

Sharing data containing personal data requires, among other things, that participants are informed about data sharing and have given consent, that access to the data is controlled, and that further use is regulated through agreements. It should also be documented who has access to the data and how it is used.

At present, there are few established solutions in Norway for sharing such datasets, particularly for qualitative data. At the same time, work is underway to develop solutions in this area, including through national initiatives related to DataverseNO.

Sharing research data that is not fully anonymized is therefore challenging, but possible. Assessments must be made in each individual case. Contact the library if you need guidance.

Research data are typically published towards the end of a project or after it has been completed.