The long-term accessibility and preservation of data is key to upholding scientific integrity and reproducibility. Data repositories are online storage systems where datasets linked to research can be deposited and accessed for a multitude of purposes. Choosing a data repository to deposit your research findings in is a choice typically up to the researcher, so how do you go about selecting one?
In this article, we’ll walk you through what a data repository is, types available to researchers, and how to choose one that’s right for you when it comes to depositing your preprint data.
What is a data repository?
For data to be shared with others, it first needs to be retrieved. And for this to be possible, the data must be stored in an easily discoverable location.
The history of data preservation goes back to the 1890s, with the creation of the index card catalog by Paul Otlet and Henri La Fontaine, an analog database serving as the first proper attempt to systematically organise all global knowledge.
Fast forward to the 1990s and the rise of the internet and early open access movement and we see the emergence of the first preprint servers such as arXiv. These online archives were the beginnings of the shift towards long-term preservation, organisation, and reuse of scholarly data.
Data repositories act as digital storage systems that centralise datasets and make them easier to discover and access. These datasets use logical means to categorise the data they contain, allowing researchers to search by terms most relevant to their needs.
The long-term preservation of data has significant benefits for society. These benefits include:
- Upholding of scientific integrity.
- Research funded by taxpayer funds and any subsequent data is available in the public domain.
- Better responses to global crises (such as pandemics, etc).
- Reduces the need for animal testing.
- Cultural and political heritage.
Types of repositories
There are various types of data repositories, each with their own uses. Let’s look at some of the most common types.
Multidisciplinary/generalist
Multidisciplinary or generalist repositories store data from across all research disciplines. This is especially useful for data that might strand various disciplinary issues. Given the broader audience, such repositories may potentially increase the visibility and reach of your data. These repositories also tend to be free for users.
Examples of multidisciplinary or generalist data repositories include Zenodo, Figshare, and Dryad.
Discipline-specific
Discipline-specific data repositories invite researchers working in a specific field to deposit their data. These repositories may focus on niche fields such as nutrigenomics or a specific type of organism.
Re3data is a research data repository you can use to search for discipline-specific repositories.
Institutional repositories
Institutional data repositories collect research data produced by members of a college, university, or other institutional body.
Funder repositories
Certain funders may expect data to be deposited in a certain repository. For example, the Economic and Social Research Council (ESRC) expect researchers to prioritise submitting their data to the UK Data Service.
How to choose a data repository
Choosing a suitable data repository depends on many factors, including your discipline, legal and ethical considerations, funder expectations, and institutional commitments.
Beyond the fundamentals that might influence your decision, there are also desirable services and specifics to look out for when making your choice. These include:
- A credible reputation within research either for multidisciplinary or discipline-specific data preservation.
- Can guarantee long-term preservation of your data, even in unforeseen circumstances.
- Uses persistent identifiers such as Digital Object Identifiers (DOIs) that ensure research remains accessible even when site architecture is changed (such as landing pages being moved to a different URL address, etc.) and that citation is easier.
- Versatile metadata that enable discovery and reuse and are tailored to the controlled vocabularies of your desired research community.
- Datasets being widely available to the public and others through application of open access licensing.
- Supporting of the FAIR Data Principles through the provision of open metadata.
As a researcher, your needs and the characteristics of your desired data repository are the driving factors influencing your decision. Still, with numerous to choose from, it can be overwhelming to decide. If you are unsure about which data repository to use, you can seek further advice from your peers or the preprint platform you are submitting your research to about which data repository might be best for you.
For researchers posting a preprint with Preprints.org, authors are encouraged to follow FAIR Data Principles and upload data to a recognised repository before posting a preprint (such as one from Re3data). Authors can also upload supplementary materials in any format along with their preprint.