The example doesn't only highlight the difference between correlation and causation, it's just one of the ways naive data fetishism is ineffective science. There is no science without experimentation, and there is no experimentation without theoretical frameworks.
My point is not that we should not test. My point is that we should, and that we can draw connected conclusions based on test results that enable us to predict outcomes in similar situations and conditions, but that which are fundamentally novel. This is why we have a theory of gravity, we can observe how physics works generally to apply to new particular instances, rather than testing it each and every time.
The example doesn't only highlight the difference between correlation and causation, it's just one of the ways naive data fetishism is ineffective science. There is no science without experimentation, and there is no experimentation without theoretical frameworks.
Very well said, much more clear than me!
My point is not that we should not test. My point is that we should, and that we can draw connected conclusions based on test results that enable us to predict outcomes in similar situations and conditions, but that which are fundamentally novel. This is why we have a theory of gravity, we can observe how physics works generally to apply to new particular instances, rather than testing it each and every time.