There are three modes of science: descriptive, explanatory, and applied. The descriptive mode consists of systematic observation of phenomena followed by discovery and empirical confirmation of general laws covering regularities in the observations. The explanatory mode uses the observations and laws with assumptions that fill in gaps to tell a story of how and why a particular sequence of events happened or something came to be. The applied mode uses the results of the descriptive and explanatory modes in developing a product or technique.
The descriptive mode of science consists of detailed descriptions of phenomena and heuristic rules that generalize descriptions of related phenomena. In the descriptive mode one strives for objectivity with thoroughly empirical science methods unadulterated by personal beliefs, unobservable entities, or unconfirmed assumptions. These scientific descriptions do not explain phenomena except in the minimal sense that they can replicate aspects of phenomena that are subsumed in generalizations. For example, a generalization about planetary motions would show that their orbits are elliptical but not why the orbits are elliptical or how planets move.
The explanatory mode of science is built on the descriptive mode by taking descriptions and generalizations of phenomena as its starting point. The scope of each generalization is taken to be as universal as possible within any limitations imposed by other generalizations. For example, if there is a generalization about a certain fish population, in the explanatory mode it becomes a law about all fish populations unless there are other generalizations that limit the scope to some fish populations.
Explanations are developed in the explanatory mode that are consistent with the descriptions and generalizations of the descriptive mode with non-empirical matters kept to a minimum.
In the explanatory mode the beliefs of the person presenting or endorsing the explanation may affect the particulars of the explanation.
In the explanatory mode thoroughgoing empiricism is an ideal to strive for but compromises may be made by positing unobserved entities or making assumptions that enable general laws to be a sufficient basis for a story that makes the how and why clear. The making of assumptions should be kept to a minimum but may be necessary to provide sufficient grounds for invoking a law. For example, some boundary conditions may be unobserved but be required for substitution in the expression of a law. While the highest standards of objectivity should be maintained, there remains a certain subjective element in whether or not a reasonable person will be satisfied with a particular explanation. A reasonable person may be remain unsatisfied and object to an unobserved entity or assumption. To continue the previous example, an explanation of planetary orbits that appeals to a force of gravity as the causal agent producing elliptical motions may be sufficient for most people but be unacceptable to some on the grounds that a force is an unobserved entity.
The applied mode takes the descriptions and explanations with perhaps non-scientific influences such as heuristics (rules of thumb) and aesthetics to design a product or technique with some purpose in mind. Engineering is the systematic practice of the applied mode. To continue the example once more, an engineer may use the law of gravity along with other laws such as aerodynamic laws to develop the path that a satellite launch must take to attain a particular orbit.
The explanatory mode has some of the characteristics of the descriptive mode and the applied mode. It is descriptive in that it includes descriptions of phenomena to be explained and it is applied in that it applies laws to construct explanatory stories of phenomena.
There is no single method to confirm a proposed scientific law but the basic idea is to increase the weight of evidence for it, especially in contrast to other possible laws. Experiments that challenge the law but lead to new observations consistent with the law increase the weight of evidence. Explanations for events or artifacts that either lacked explanations or had unsatisfactory explanations because of the unobserved entities or assumptions increase the weight. Useful applications developed using the law that would not likely have been developed without it increase the weight. Thus we see all of the scientific modes at work in confirmation.