As a professional scientist, I am accustomed to observing some phenomenon that is interesting to me, wondering about its significance, looking at background data (if any exist, and yes, “data” is the plural form of “datum”), and then trying to figure out the principles behind.
That works well, for the most part, in systems that can be controlled and only one variable at a time changed. That is the essence of the scientific method as it generally practiced. I dispensed with the classical steps, which are observation, formulation of a hypothesis, experimentally testing the hypothesis, refining or rejecting the hypothesis, and then doing more and better experimental tests to test the hypothesis further. If one is very lucky, the refined hypothesis becomes a theory, and if no exception after crushing peer review, becomes a principle or a law.