Stated very simply, data sharing is the release of data for use by others.
There are a variety of models for how data may be shared, including:
When sharing, your data should not only be available but also usable. Making sure your data is well organized and documented as you are working on it will go a long way towards ensuring its usability when it comes time to share. One term you may encounter in this context is FAIR, which is defined below.
To be FAIR, your data should be:
|F||Findable - The first step in (re)using data is to find them. Metadata and data should be easy to find for both humans and computers.|
|A||Accessible - Once a potential user finds the required data, they need to know how they can be accessed.|
|I||Interoperable - The data need to interoperate with applications or workflows for analysis, storage, and processing.|
|R||Reusable - Data should be well-described so that they can be replicated and/or combined in different settings.|
Following the practices outlined in this guide will help you ensure that your data meets these guiding principles.
How and where you share your data will depend on the characteristics of your data. Datasets that are especially large or contain sensitive information can not be shared in the same way as datasets that are (comparably) smaller and are associated with less risk. A good rule of thumb when sharing data is to put it somewhere where it will be found by other researchers. For publicly shared data, this may involve choosing a particular repository. For data that is more restricted, this may involve giving precise instructions on how qualified researchers can gain access.
The figure below is designed to help you navigate the ever-changing data repository landscape. The registry of research repositories (re3data) is an extremely helpful resource for identifying repositories that are specialized for certain types of data.
Making data and other materials available “upon request” means that the requester must contact a member of the research team (often a corresponding author) and a team member must have the data on-hand (in a usable format) in order to respond to the request.
Over time- as contact information changes, team members move on, and data is archived- both requesting data and responding to requests become more difficult.
Supplementary materials are an important part of the scholarly record, but it is not uncommon for links between them and the articles they are associated with to break down. Whenever feasible, we recommend uploading data into a repository that is designed to preserve and make data accessible to others and then link to and/or citing that dataset in your manuscript.
The Dryad Digital Repository is a curated resource that makes research data discoverable, reusable, and citable. Dryad provides a home for a wide range of data types and is free to use for all Stanford affiliated researchers.
Key features of Dryad:
See their FAQ page for additional information about Dryad's features.
There are a variety of models and potential platforms for sharing your datasets with other researchers. Lane Library recommends Dryad as a way to openly share datasets that do not fit into more specialized repositories. For more information about Dryad, contact your liaison librarian.
Dryad uses ORCID iDs for login. The first time you log in, you will be asked if you are affiliated with a member institution. After selecting Stanford from the drop-down menu, you will be asked to sign in using your Stanford credentials. On every subsequent login, you will only have to use your iDs.
Once you have logged into Dryad, you can begin the process of publishing and sharing your data. After clicking Start New Dataset, you will be prompted to begin entering metadata. Good metadata (also called data documentation) is vital for ensuring that your dataset can be discovered, understood, and used by other researchers.
Dryad only requires that you complete the title, authors, and abstract fields, but we strongly recommend that you complete every field and upload additional documentation (e.g. data dictionaries, readme files, etc) alongside your dataset.
Dryad has two different methods for uploading data. Both methods allow you to upload multiple files.
Once you've uploaded your files, you can decide to submit them to the curation process immediately or keep them temporarily private for peer review. During the curation process, expert curators perform basic checks to ensure that the title and abstract are meaningful, there are sufficient methods and usage notes, that files can be opened, and that no sensitive information of material subject to copyright restrictions have been inadvertently included in the dataset. As an author, you can review the curation process for your dataset.
If you are plan to use Dryad to publish and share your data, please feel free to use or adapt the following description when completing data management plans or other documents:
Stanford University is a Dryad member institution. Dryad is an open source tool for data publication and digital preservation. Datasets deposited into Dryad are permanently archived in a CoreTrustSeal-certified repository. Data files are regularly audited to ensure fixity and authenticity and are replicated with multiple copies in multiple geographic locations. Professional curators examine all Dryad deposits to ensure the validity of the data, apply robust metadata, and make certain that highly sensitive information has not been inadvertently included. Datasets deposited in Dryad are automatically assigned a Digital Object Identifier (DOI) and are indexed by Google Dataset Search and other tools to enhance discoverability.
More information about Dryad's features, see this page. For additional assistance in describing Dryad or to discuss how it can be integrated into your research workflow, contact your liaison librarian.
Increasingly, there is an expectation that researchers will share their data. Data sharing can be a complex endeavor and, though we think very highly of Dryad, Lane Library recommends that you choose the method for sharing that is right for you and your data. Answering the questions below will help guide you through this process. For additional assistance, please see our upcoming classes and events page for workshops related to data management and sharing or contact your liaison librarian.
In some cases, your research funder or the journal publishing your work will specify that your data should be shared through a specific repository. For example, some projects funded by the National Institute of Mental Health are expected to share their data through NIMH Data Archive. In cases like this, we recommend that you share your data through the required repository.
Please note that some requirements state that data should be shared, but do not specify where. In such cases, refer to the next question.
If your research community typically shares the type of data you are looking to share through a specific repository, we generally recommend that you use that repository. To find repositories specialized for particular types of data, we recommend searching the Registry of Research Data Repositories (Re3Data).
If there is not a repository that is specific to the type of data your working with or if you have other concerns about sharing your data, see the next question.
Certain characteristics of your data may determine how and where it can be shared. For example, if you are working with big data (over 300 GB) or data that contains personally identifying information, we recommend scheduling a consultation with your liaison librarian so we can refer you to the appropriate group on campus to help you determine your options for making your data available.
However, if you are simply looking for a general-purpose data repository, we strongly recommend Dryad. Stanford Libraries also maintains the Stanford Digital Repository (SDR) which is recommended to Stanford University affiliates.
Data are generally considered to be citable products of research, meaning you should cite them when you use them and look for repositories that facilitate easy citation.
Digital Object Identifier (DOI) - A unique alphanumeric string to identify content and provide a persistent link to its location on the internet. DOIs are commonly assigned to journal articles (i.e. https://doi.org/10.3897/rio.4.e26439) but can also be assigned to datasets. Getting a DOI for your dataset helps ensure that other researchers will be able to find and use it.
Accession Number - A unique number (or alphanumeric string) assigned by a database as a means of locating a specific object. When citing or pointing to a dataset, you should at least provide its accession number.
Research Resource Identifiers (RRID) - Unique ID numbers assigned to help researchers cite key resources (antibodies, model organisms, software projects, etc) in the biomedical literature. RRIDs are not applied to datasets directly but should be included in related documentation.
ORCID iD - An alphanumeric code that uniquely identifies scientific and other academic authors and contributors. Lane Library recommends that every researcher claim their ORCID iD. For more on registering and using an ORCID iD, head over to our dedicated guide page.