Including Data Sharing & Management Costs in an NIH Grant Proposal

The National Institutes of Health (NIH) encourages researchers to include data management and sharing (DMS) costs in their grant proposals. This guide summarizes key points and offers SPARC-specific information.

What costs are allowed?

NIH allowable costs associated with:

• Curating data
• Developing supporting information
• Formatting data according to accepted standards
• De-identifying data
• Preparing metadata
• Local data management (before submitting to SPARC)
• Preserving and sharing through established repositories (including SPARC)
  Note: These costs must be incurred during the grant period, even if some activities happen after the grant ends.

NIH links:

• Data Management and Sharing Policy
• NOT-OD-21-015

Where do I document my Data Management and Sharing budget requests?

1. Within the Data Management and Sharing Plan. Explain how your scientific data and accompanying documentation will be managed and shared.
2. Within the Budget. Include each item in the appropriate category.
3. Within the Budget Justification, in a section labeled “Data Management and Sharing Justification”. For a modular budget, this would be in the Additional Narrative Justification attachment; for a detailed budget, this would be in the Budget Justification attachment.
   The justification should include a brief summary of the type and amount of scientific data to be preserved and shared, and the name(s) of the established repository(ies) to be used. It should also indicate general cost categories (such as curating data, etc.); each category should include an amount and a brief explanation.

📘

Here is a SPARC-specific Data Management and Sharing template you can use to get started

The cost of your DMS plan depends on various factors. Consider the following:

General Advice:

• Plan in advance! Just as manuscript submission requires time to finalize, so will planning your data sharing costs.
• Include the time and effort involved in documenting your data collection. SPARC datasets must be accompanied by experimental protocols that explain how the data were obtained (we recommend protocols.io).
• Overestimate costs, especially people’s time. Or simply include a cushion.
• Don’t forget the costs of organization and curation.
• Establishing good data management practices in the lab will save you time and money down the line.
• The advantage of planning ahead is that you’ll be able to streamline data collection, management, storage, and sharing.

Additional Advice Links:

• Open Data Commons Estimating Costs for Data Management and Sharing
• Estimating Costs for Sharing Data Through Open Data Commons: Curation and Data Preparation
• NDA Data Contribution
• UCSF PI Budgeting Guidance
• Utrecht University Research Data Management Support

SPARC costs

There is a standard procedure for submitting SPARC datasets. They are required to follow the SPARC Data Structure (SDS), a standard for organizing and naming your data files. Furthermore, SPARC also requires your data to be accompanied by an experimental protocol publication with a corresponding Digital Object Identifier (DOI). Our SODA software will help you through the process.

SPARC can help with data management and sharing services, as negotiated (contact us). Basic curation services are provided, but we can also negotiate professional assistance. Finally, we can assist with grantsmanship, DMS plan preparation and budgeting.

SPARC Links:

• Submission Walk-through Intro
• What We Offer

Internal costs

In addition to the straightforward costs associated with a SPARC submission, you may wish to budget for internal costs. For example, you may wish to budget for people within your lab to help prepare the data, or hire a data management expert. You should budget time and resources for writing the required experimental protocol, as described here. The start of a project is also an excellent time to institute data management practices according to your own standards. Below are some questions to ask yourself as you consider the resources you will need to execute your DMS plan:

• What shape will my data be in (how “tidy” is it?) and what will it take (resources and personnel) to get it into condition for submission to SPARC?
• What documentation (a “data dictionary”) is needed on our end to make any custom elements useful to others?
• What lab data management practices and protocols should I set up to make sure that data collection is done correctly? What will it take (again, in resources and personnel) to create and maintain these practices?
• Do people in my lab have the expertise to do the above items, or do I need to hire an expert?

Links:

• Recommended research data management practices
• SPARC checklist of factors to consider

Information needed to make the calculation:

• Number of subjects
• Number of sites
• Number of times data submitted
• Number of data structures
• Number of unique experiments
• Number of publications expects
• Hourly pay for PI
• Hourly pay for data manager
• Estimated hours of work for PI
• Estimated hours of work for data manager

Other costs:

• Data storage
• Data transfer
• Data preservation and sharing
• Data cleaning
• Digitizing any analog data
• Converting to sustainable file formats
• Converting to SPARC formats
• Charge for data repository itself
• Hiring expertise
• Consider a data manager