What does it mean that schemas must be tested against reality?

An untested schema is a hypothesis not knowledge.

Example: You believe that morning meetings kill your team's productivity. You've held this schema for two years, cited it in every process discussion, and reorganized your calendar around it. But you have never actually measured whether your team ships more when mornings are meeting-free versus when they are not. Your schema feels like knowledge — it has conviction, consistency, and emotional weight behind it. But without a single test against observable outcomes, it remains a hypothesis dressed in the authority of repetition.

Try this: Select one schema you hold with high confidence — a belief about how something works in your life, your team, or your field. Write it as a falsifiable claim: 'I believe [X] because [Y], and if [Z] happened, it would prove me wrong.' Then identify one observable test you could run in the next seven days. The test does not need to be elaborate. It needs to be honest — genuinely capable of producing a result that contradicts your belief. Write down what result would confirm your schema and what result would challenge it, before you run the test.