The numbers are self-reported by the person who made the decision. Harvard case study just means it's interesting enough to study, not that it's replicable or that the causation is clean

→ View original post on X — @aihighlight