Carnap = see instances and use induction to create hypothesis
Everything else they say about the scientific method is the same.
Herbert Simon, in thinking of sciences of the artificial, used the Carnap method to create a ton of valuable science.
It's not that post-facto is bad - carnap showed otherwise.
It's about the quality of the science and the specificity of the recommendations, that lead to bs.
For instance you can mitigate for survivorship bias by studying the dead as well as the living, and driving deep into the differences. But that's a lot of work!
As a theory, effectuation studies the difference in decision making processes between expert entrepreneurs and corporate CEOs. The result, while similarly post-facto, is delightful. See effectuation.org - Vinod Khosla remarked it was the first useful study on entrepreneurship he had ever read.
Tally rocks! Thank you.