[ 6/4/20 ]
While I agree with Marty Fouts, that science done in reality is vastly more complex than anything taught in secondary school or undergrad classes, there is a general theme in the method that is important, and fundamental, and to which everyone returns at some point.
That fundamental thing about science, is letting reality (whatever it is) be the final arbiter on questions (rather than any form of Truth or dogma).
And real science deals in probabilities in all cases, and in practice all scientists will have some set of things sufficiently well tested that they don’t think about uncertainties, they treat them as truths in practice.
So the general form being, come up with some set of hypotheses to explain something observed, design an experiment that should give a reasonable probability of distinguishing between them as to which is more useful, then carry out that experiment.
And it can get extremely complex, as different individuals can bring fundamentally different sets of interpretive schema to an issue.
A good scientist will acknowledge that all models are probably wrong, but some are vastly less wrong than others.
It’s most often useful to be using the least wrong model available.
And it is more complex than that even, as different models have different computational costs (thence time required), and if there are deadlines approaching, then simplifications must be made in order reach a conclusion. That problem is potentially infinitely recursive through the space of all possible models and all possible logics, and causes a vast number of real issues in society today (politicians are usually working to deadlines, and usually like simplicity – even where it does not exist).
So anyone beginning the journey into science has to start with relatively simple models. Some people like to hold onto such simplicity. Exactly where one sets the boundaries on that is a major determinant on how far one gets on any particular issue. It is easy to get lost in complexity.
There is usually a fair degree of pragmatism and a fair number of useful heuristics employed in any practical scientific endeavour.