Notably, these criteria for exclusion focus on the status of the dataset (e.g., is it public? does it already exist?), not the content of the dataset nor what will be done with the dataset, which are more accurate criteria for determining the risk profile of the proposed research.