PHY.2:4.5 - Obtain a consequence and determine its reach
Manipulate the source according to its proposed laws. Start with a small input, a change of one interaction, or a limiting case that exposes how the construction works. Obtain the consequence by a derivation, a qualitative physical argument, a computation or an experiment. Use a known limiting case to find an omitted interaction or a sign error where it can discriminate between constructions.
Trace the obtained consequence back through its physical and mathematical premises. A computation can determine a property of the proposed source model. Interpreting an apparatus reading adds the account of the apparatus. Applying either result to a physical target adds the target correspondence and the physical premises used there. An existing explanation can supply those premises; no new evidence collection follows merely from enumerating them.
If the target mechanism remains proposed, use the result as a consequence of that hypothesis. Ask what change would make competing mechanisms disagree. This can reveal a useful intervention, a different observation interval or a parameter relation to examine. MMP.7 helps when the proposed distinction must be expressed through recorded data; C.11.DUA determines whether obtaining that distinction is worth the effort.
Different sources can contribute complementary results. One may expose an invariant or a bound, another the effect of noise or a difficult regime. Compare them for the usefulness, range and cost of the required contribution. G.5 distinguishes alternatives retained for later choice, non-dominated candidates and a set whose members contribute together. E.23 supports repeated improvement when a stated evaluation can judge the changed source.