In, HyBIST baseline is defined as a Cartesian product of ‘What-to-test’ and ‘Test-for-what’.
What-to test: In real life usage a typical user uses a combination of user stories to get his/her job done. So we need to test not only a user story but a collection of user stories that form a flow. So “what to test” is and individual user story and a collection of user stories.
Test –for-what: In agile context, this is called ‘Acceptance Criteria’, a set of user story that a user story is expected to satisfy to deliver value to the user. Acceptance criteria is not limited to functionality alone, it may include non-functional attributes also.
Connecting these two, Baseline is about clearly understanding the user stories (&combinations) and various types of tests to be done, to meet the various acceptance criteria.