Well, one technique that works pretty well along these lines is reporting detailed experimental results demonstrating (or failing to demonstrate) the principle one wants to communicate/understand, and encouraging one's peers to reproduce the experiments.

Not quite as good, but sometimes more accessible, is selecting some theoretical examples of the principle one wants to demonstrate on the basis of a general guideline (rather than a guideline chosen case-by-case so as to return preselected examples) and working one's way rigorously through those examples to see where they lead.

The How to Change Your Mind sequence isn't a bad starting point.