Data Cards are very comprehensive and describe many facets of the dataset, some of which may not always be readily available. Look for new knowledge that becomes available as you fill out a Data Card.
In completing a Data Card, teams may find the need to perform additional analysis to complete the Data Card – or discover limitations in the dataset that can prompt additional data collection.
Track answers that you’re uncertain of.
A partially irrelevant or unanswered question can be indicative of gaps in knowledge about the dataset. These can be used to assess new analyses or documents that can become available over time. As your dataset goes into circulation, keep an eye out for these kinds of analyses and documents from across the dataset lifecycle. Revisit your Data Card every few months with a specific agenda to complete these questions.
Analyses that are rarely one-and-done.
It’s easy to assume an answer is sufficient once you have captured the results of analyses that are already available. Don’t stop there – if your Data Card is missing evidence for an answer, consider what analysis needs to be done, or what additional data might be needed to answer those questions?