Research Data Service

Structure your data

The structure of data in the DataVault

You will need to decide how you are going to structure your dataset(s) in the DataVault, i.e. which subsets of your data, if any, will require separate vaults.

A vault represents one dataset and is associated with one Pure dataset record, i.e. it is described by the information (the metadata) in the Pure record. For example one 'experiment' or project would have a dedicated vault, and would be described by one Pure record listing the funder(s) of the work. Within one vault you may have multiple deposits, but they will all relate to the one research question. All the contents of a vault will be reviewed for deletion or retention at the same time. However, individual deposits may be deleted while others are kept.

A vault can contain many deposits, deposited at different times.

When a vault is created, Research Data Support staff will check to ensure the Owner has appropriate resources in place.

A Principal Investigator might, for example, create multiple vaults to archive data relating to different projects. This would be necessary if they wanted to provide access to different users - access and roles are assigned per vault. 

Diagram of DataVault structure showing vaults, metadata records and deposits
Example structure for data arising from Project A and Project B