How do you handle conflicts or disagreements with team members regarding data interpretation when the evidence is ambiguous?