Comparison rule: Use change over time to choose better follow-up observations. Do not treat correlation as a diagnosis.
Compare two locally saved Notebook entries to see what stayed stable and what changed. A changed measurement and a changed symptom can be related without proving that one caused the other.