Develop clear guidance on proprietary datasets

[sergegoussev]

Goal of info to include

• Most researchers will either use their own NSO datasets or other proprietary datasets such as Nielsen. We need guidance on the topic
• Specifically, there are two scenarios:
  • If NSOs apply known methods on internal datasets, the guidance is to aim for reproducibility within their NSO and at most replicability externally (as sharing the input and likely output data is not necessary)
  • For new methods, use of openly available (versus proprietary ones) should be recommended. The alternative is replicability if that is not possible

Success Criteria

• When options for propriety and internal datasets are clearly outlined in the guidance