Put the question first
When an experiment starts, its purpose should be visible. Keep the hypothesis close to the configuration so it stays part of the work rather than a note in another tool.
Make the transitions clear
Moving from an idea to an experiment and then to evidence should feel like following one thread. Clear states and consistent language help people understand where they are.
Leave room for judgment
A workspace can organise evidence and make comparisons easier. The researcher still needs room to question the assumptions and choose what to investigate next.
The interface should make the research easier to understand.