UNC Dataverse Admin and Curators can delete draft datasets and dataverses. Draft datasets must be deleted first before you can delete a draft dataverse.
If you have already published your datasets, please see our section below on Deaccessioning a Published Dataset.
Delete a Draft Dataverse
A draft dataverse can be deleted so long as there are no published or draft dataset records within it. To delete a draft dataverse, navigate to the project dataverse and click on ‘Edit’, then select ‘Delete Dataverse’.
Delete a Draft Dataset
There may be times when you need to delete a draft dataset. Since it has not been published yet, Dataverse has not registered its DOI (persistent identifier), which means you can simply delete the draft from your project dataverse.
If you have published your dataset already, please see the Deaccession Published Datasets section of this guide for instructions.
To delete a draft dataset, click on the dataset record and then select ‘Edit Dataset’. Click on ‘Delete Dataset’ from the drop-down menu.
A pop-up window will appear asking if you are certain that you want to delete the draft. Deleting the draft will remove it entirely from UNC Dataverse.
A deleted draft cannot be recovered, so make sure this is what you want to do!
Click on ‘Continue’ to permanently delete your draft dataset. You will see a green success bar at the top of your project dataverse alerting you that the draft dataset has been successfully deleted.
Deaccession Published Datasets
Deaccessioning is the term used to describe ‘deleting’ a published dataset from UNC Dataverse. UNC Dataverse does not delete the entire dataset. As a data repository, we are committed to transparency and preservation of our holdings. This means that when you deaccession a dataset record within UNC Dataverse, the persistent identifier (DOI) will still resolve to the dataset record; however, the data files and most of the metadata will be removed. Deaccessioning should only be completed if the data are in violation of UNC Dataverse Terms of Use (data contain PII or PHI, copyright violations, takedown requests, etc.).
With a deaccessioned dataset, users will be taken to a ‘tombstone’ dataset record which provides information about when the data were removed and why. If the data are available in another location, instructions for accessing those data will be included in the deaccessioned dataset record. This ensures that users can track the version history of the data and can reach out to the data producers if the data are available for download outside of UNC Dataverse.
Do not deaccession a record on your own.
If you need to deaccession your published dataset, please Submit a Help Ticket, and provide a link to the record in question. We will review the record and work with you to find the most appropriate solution.