Intended audience: DataKind Volunteers
At the end of every project, we need to make sure that we have a plan in place for appropriately taking care of the partner organization’s data. The decision for data publication or destruction should be made in partnership with the partner organization during the Design Stage, as part of the data management plan. Once the project is completed, review the policy selections and any wrap-up steps in the data management plan. Confirm the appropriate wrap-up steps have been completed and update the data management plan to indicate who completed them and when. For more on responsible data destruction and retention, check out The Engine Room’s “Becoming RAD” resource.
Data Destruction
If the expectation is to destroy the data, make sure to confirm with all the project volunteers and staff that they have double checked that they did not accidentally copy the data anywhere on their personal computers. Additionally, double check all Google Drive folders, to make sure the data wasn’t copied into any other folders. For example, sometimes data can be found in the “Design” folder from the data audit or the “Share” folder for sharing project results. Additionally, complete another code review to ensure that data isn’t copied anywhere in the code on GitHub. Carefully delete both the current version and any version history of anywhere that data is found. Make sure that anyone who deletes data on their personal drives also clears their trash.
Data Publication
If the decision is to publish the data, make sure that the final dataset is cleaned, checked, and properly labeled and formatted with a descriptive and clear data dictionary. Work closely with the Project Champion in reviewing the data and the associated data documentation before publication. Although this should already be the case, double check that there is no possible personally identifiable information or variables that can together substantially limit the number of people that a row might refer to to the point of potentially putting anyone’s personal information at risk. Confirm the dataset is ready and the plan is appropriate with all stakeholders at the partner organization and DataKind before publishing the data.
Contributer(s): Benjamin Kinsella, Rachel Wells
Contact us
If you would like to learn more about us, partner with us, or get in touch, visit our website or email community@datakind.org.
Subscribe