Stopping an answer when evidence is insufficient

This anonymized case study is currently maintained in Japanese. It explains how an answer path was changed to stop when required evidence was missing, while keeping unmeasured business outcomes explicitly unverified.