Research Data Management Guidelines

(2018-04-11, Ayan Segupta & Ari Lingeswaran)

You should describe in your research data management plan how you will archive and share (if required by your funders) your data; and how you finance your data management. You can generate Data Management Plan and find further useful information on DMPOnline. It has several templates for most UK, EU and US funders, as well as institutions; together with corresponding guidance. You might also find information regarding durations, intended audience, etc. there.

You can find further information and an institutional guide on how to write Data Management Plan, you can find here.

Research data can be archived on the Royal Holloway Figshare Data Archive. Arkivum provides the storage for data deposited there. Free storage space available to individual users, unfunded projects and funded projects is currently set at up to 1GB, 5GB and 0.5TB, respectively. Costs for storage of data more than 0.5TB is according to the Arkivum Pricing, and the current (11th April, 2018) price per TB per year ranges between £52 (data > 1000TB, for 25 year paid upfront) and £360 (data < 100TB, pay as you go).

It is also best to try to recover storage costs as a direct cost to the project. This would be especially helpful for project data exceeding the 0.5 TB allocation.

You can find further information and step-by-step descriptions here.

Check out discipline-specific repositories, such as OpenNEURO for raw data (successor of openfMRI), neurovault for statistical maps, which are typically free, but may require specific formats Brain Imaging Data Structure (BIDS), NIDM Results). Tibor is happy to help to put your data in the required format.

BIDS is strongly recommended for both OpenNEURO and internal (MRIRaw on the cluster / FigShare) for several reasons:

  • It is getting to become a widely accepted and recognised standard
  • There are tools, such as automatic analysis and other BIDS Apps also available on Open NEURO, that supports BIDS creation and processing especially for standard use cases
  • It is economic (single compressed 4D file for functional and diffusion-weighted images)