Systems analysis is a methodology applied to problems in systems in which specialists may combine with decision-makers to blend in a concerted way the diverse approaches available. Thus each solution arrived at in this way is a work of reasoned art, not the result of a prescribed methodology or formula that applies to all cases.
The major characteristics of systems analysis are that:
1. It is a powerful technique for grasping ill-structured, large, complex problems of choice under uncertainty (more effectively than if its individual parts were examined in isolation). It looks at the broad goals to be achieved and examines the costs, effectiveness, and risks of the various alternative approaches to achieving the objectives;
2. It permits and encourages the judgement and knowledge of experts to be joined in a systematic and efficient manner, and facilitates the blending of the judgement of generalists and managers with the expertise of specialists;
3. It employs the scientific method, namely: (a) it is open, explicit and results can be verified by others, even though quantitative and qualitative information are mixed; (b) analysis is systematic and objective; (c) hypotheses are tested and verified by appropriate methods; (d) information is quantified wherever possible;
4. It constructs and operates within models or simplified abstractions of the real system situation appropriate to the problem;
5. It evaluates alternatives by a careful assessment of costs against benefits, making cost-benefit analysis an important part of systems analysis when applicable;
6. It deals with practical problems, and not theoretical problems;
7. It attempts to deal explicitly with uncertainty;
8. The context is often broad and the environmental i.e. external factors are usually very complex, such that simple problems exceptionally come within its scope;
9. There is an absence of a universal guiding theory, making systems analysis a methodological art rather than a science, one that seeks out the theory or analytics that are appropriate in each case;
10. It is focused more on exploring the implications of alternative assumptions than on analysing in extensive detail the implications of a single set of assumptions;
11. It is ordinarily not concerned with computing an optimum solution but rather with giving the decision-maker a range of choices and outcomes. It emphasizes design of new solutions and widening the range of alternatives, rather than selecting the best alternative from a predetermined range;
12. It lacks definition as a methodology, thus the foregoing characteristics are negated by some authorities, to wit: it applies to all systems, simple or complex; it has a distinction methodology which applies in all cases; it is the domain of specialists only; it is not always scientific in that it accepts some unverified hypotheses of systems theory; it can be rigorously applied to exhaustive analysis of a single solution, etc.
The organization of a system is simple if the system is a serial or an additive complex of components, each of which is understood. As soon as strict sequential sequences or linear additivity is transcended, an organized system becomes rapidly more complex, usually too complex for detailed analysis into superposable parts or effects. At the other extreme from organized simplicity is chaotic complexity where the number of entities involved is so vast that the interactions can be described in terms of continuously distributed quantities or gradients, and do not need to be specifically identified with regard to the individual entities. Such systems can be described by the methods of statistical mechanics which merge with those of classical mechanics when the collections of entities are treated as continuous.