Data Management Planning

Research Data Management includes a wide range of elements in a research project: from how data will be collected and stored, how access to it will be managed during the research period, and how that data will be preserved and made reusable after the study is concluded.  During project design these elements of a study may be formally documented in a Data Management Plan (DMP) submitted as part of a grant application. Even when a formal document is not required we strongly encourage the creation of a detailed plan for managing research data over the full lifecycle of a project as this planning process can bring multiple benefits. 

For more about writing a formal DMP, see our "Creating a data management plan (DMP)" article.

Potential benefits

Some examples of potential benefits that data management planning can bring to your research include more accurate metadata collections and the anonymization of sensitive data. Selecting metadata standards for your data at the beginning of a project can help to reveal what information about that data needs to be recorded at the time of collection or processing when that information can be captured most easily and completely.  These best practices help improve the accuracy and completeness of the data you are collecting while making that data easier to work with and reuse. Similarly, implementing anonymization and minimization strategies early in the data management workflow can reduce the potential for accidental disclosure of legally or ethically protected data while increasing the portion of your workflow that can be publicly documented. Additional examples of best practices in data management as they relate to a wide variety of different research contexts are available from the DataOne project’s best practices database. 

Using planning to inform technology decisions

Once the key elements in a data and research management workflow for a study have been identified, it becomes easier to select appropriate tools to enable and support those steps. Recognizing your organizational system and software needs early enables you to avoid the uncertainty and inaccuracies that can creep in when research data needs to be moved around in unplanned ways. Routine requirements such as providing data access to collaborators at different institutions, transitioning ownership of data from a colleague departing the research team, and providing anonymized access to materials during peer review might all lead to an unexpected need to copy or migrate data to new systems if not considered in advance. When data needs to be moved or processed in ways not addressed in your data management planning, confusion is likely to arise as versioning conflicts or lost metadata (e.g. code books that describe data labels) and may lead to errors or an increase in documentation work when you go to share your research outputs publicly. 

Considerations:

  • Ensure that all necessary stages of the research are covered by appropriate technology.

  • Plan for flexibility to handle unexpected tasks so that procedures are in place for migrating the data easily between different tools and stages of research.

  • Use version control or other provenance management tools.

  • Know where the data, analytical code, and other digital materials are likely to be stored at the end of the project. 

  • Consider who you wish to use these research outputs and what information they would need in order to locate and make effective use of them.

Using an OSF project for data management planning

OSF provides controlled access sharing and persistent online storage, has built-in tools for version control, allows for distinct, structured workspaces, and enables each section of a project to be made public or private independently. These core OSF features enable researchers to build a workspace that is the embodiment of their data management plan. Each step below is supplemented by relevant sections of the OSF help documentation where additional information may be found. Just look for the "OSF Guide" links next to the headings below.

  1. Create an account (OSF Guide)

Visit https://osf.io/register/ where you can login through your existing ORCiD account, through an affiliated OSF Institution, or provide an email address and password. Accounts are free.

  1. Create a project (OSF Guide)

When you log into OSF you will be brought to your Dashboard where you can create a new project. Create a project and give it a descriptive title to reflect the research you will be pursuing. Once your project is created, go to it by clicking the “Go to project” button that appears or by clicking on the project’s name on your Dashboard.

  1. Use components to build your project into a structured workspace that fits the data management needs identified in your DMP (OSF Guide)

Components are the fundamental unit of organization in OSF. Each component is given a unique url, called a Globally Unique Identifier (GUID), each can be made public or private independently of any other portion of the project, and each may have a different set of user permissions. Create as many components as appropriate for your planned data management workflows. At a minimum we suggest creating components for each class of research output your project will generate (Data, Materials, Analysis Scripts, etc). The “Components” section of your project page contains the “Add Component” button and displays all existing components at the top level of your project structure. 

You can also create components inside other components. This enables the creation of multi-level compartmentalized structures for organizing your project and managing access to each section.

  1. Incorporate documentation into your project structure (wiki)

OSF provides multiple ways to add documentation to each area of your project.  Each project and each component within it has a title, a description, and a wiki page that can be freely edited by project contributors. All of these can be appropriate locations to include information about your planned processes for key elements identified in the project’s DMP such as metadata collection, documentation updating, and the eventual publication of materials. 

Adding these process reminders and implementation details will help ensure that they remain prominently visible and readily accessible to the research team. 

  1. Configure access permissions for the research team (OSF Guide)

All OSF projects and components are private by default so if you wish to work collaboratively you will need to add the other members of your research team to your project by searching for their user accounts using the “Contributors” tool. 

OSF gives you substantial control when setting up user permissions. Each contributor can be given one of three levels of permission (read, read+write, or administrator), each component can have a separate list of users, and those users can be granted different levels of permissions between different components or between components and the top level project. This makes it simple to give a team member permissions on the “Data” and “Materials” components but not the “Manuscript” or “Grant Management” ones. We recommend following the principle of least permission by assigning collaborators the minimum permissions needed to accomplish their research tasks in each section of the project. In particular you should give careful consideration to who is given Administrator permission to any portion of the project. 

Edit the Wiki

Considerations when using OSF

Where possible, ensure that there are two Administrator users for each component in the project to ensure continuity should one leave the research team.

The second administrative user could be a departmental administrator or other colleague uninvolved with the research project, assuming data protection and blinding considerations allow it. In such a case, the user not contributing to the research should likely be added as a “Non-bibliographic” contributor so that their name does not appear publicly should you make some or all of the project public.

Document the procedure that should be followed when members of the research team depart, especially if you intend to make portions of the project public later. 

To help ensure that all contributors receive credit for work made public on the platform, OSF generates an automatic citation for each project and component based on the list of contributors to that section. Removing someone from the list of contributors therefore has the dual effect of ending their access to any materials while also removing their name from the citations for each of those project areas. If members of the research team leave a project, due to shifting responsibilities, jobs, etc, and their access to the research materials needs to be completely removed for the remainder of the project, you may wish to consider either manually editing the citation for those project areas to preserve their name or making a note to re-add the departing members as “read-only” contributors to those sections when you make them public. The project or component’s activity log is an invaluable resource for reviewing the contributions of each research team member.

Resources