外文资料翻译The Methodology of Systems AnalysisFor a system analysis to be undertaken,someone must think there is a problem-or at least recognize the possibility that a problem exists.That is,he must be dissatisfied with the current or anticipated state of affairs and want help in discovering how to bring about a change for the better.Systems analysis can almost always provide help,even if it does no more than turn up relevant information or indicate that certain actions offer little hope of bringing about improvement.In most circumstances,the analysis may even discover a course of action that will bring about the desired change,a course that can be recognized as the advantageous and implemented by those with authority to act.Systems analysis can also be used to present factual arguments and reliable information to help win acceptance for a proposed couse of action.In addition,it can help to prevent the chosen course from being rendered ineffective by adverse interests,misinterpretations,or unanticipated problems.Although a system analysis may be carried out without a specific user or set of users in mind,such work is not likely to have much influence other than to inform public debate.Decisionmaker is assumed to be an individual who wants to make decisions as rationally as possible by taking into consideration the probable consequences of each available course of action-selecting the "best"action by balancing its cost against the extent to which it helps to achieve his objectives and possible other benefits.The analyst's basic procedure is to determine what the decisionmaker wants,search out his feasible alternatives,work out the consequences that would follow the decision to adopt each of the alternatives,and then,either rank the alternatives in terms of their consequences according to criteria specified by the decisionmaker,or present the alternatives with their consequences to the decisionmaker for ranking and choice.In reality,the decisionmaking situation is rarely so uncomplicated.Theperson for whom a study is done is ususlly but one of many participants in a decisionmaking process,and he must use the results of the analysis as evidence and argument to bring the others to his point of view before acting.In some cases the decisions cannot be separated from the managerial,organizational,and political situation in which they are made,and the model we are assuming for the decisionmaker(called the rational actor model,or model I by Allison)must be supplemented or modified by bringing in organizational and political considerations.Nevertheless,as Allison remarks,"For solving problems.a model I-style analysis provides the best first cut. Indeed,for analyzing alternatives and distinguishing the preferred proposal,there is no clear alternative to this basic framework."This is the basic,unsophisticated view of the decisionmaking situation.As an example to illustrate the basic procedure,assuming that a legislative committee wants to propose legislation to increase highway safety,it is willing to consider three alternatives:a requirement for devices to make the use of seat belts automatic,lowering the maximum speed limit and enforcing it more strictly,and establishing higher standards for issuing driver's licenses.They ask the legislative analyst to carry out a systems analysis.It is useful to consider aproblem of thes type in terms of these elements: Objectives——what the decisionmaker desires to achieve.In the example,the objective is the problem,the alternatives may be policies,strategies,designs,actions,or whatever it appears might attain the objective.In the example,although the alternatives are limited to three types, within each type there are many possibilities to consider.Consequences——the results that would ensure were the alternatives to be adopted and put into effect.In the highway safety example,if the alternative of a lower maximum speed limit with stricter enforcement were implemented,a positive consequence(abenefit)would be a lower rate of fatal highway accidents;a negative consequence(a cost)would be the need for more police officers to be hired or taken from other tasks.Criteria——rules or standards that specify in terms of consequences(orsome subset of them)how the alternatives are to be ranked in order of desirability.For example,a possible criterion might be to rank the alternatives in decreasing order of the ratio of the reduction in the annual number of fatalities from implementing the alternative to the expenditure of public funds required.Model——an abstraction,a set of assumptions about some aspect of the world,either real or imaginary,intended to clarify our view of an object,process,or problem by retaining only characteristics essential to the purpose we have in mind.It is a simplified conceptual or physical image that may be used to investigate the behavior of a system or the result of an action without altering the system or taking the action.A model is made up of factors relevant to the problem and the relations among them essential to the purpose in mind.A model may take many forms.some common types are a set of tables,a series of mathematical equations,a computer program,or merely a mental image of the situation held by someone contemplating an action.In most systems analyses explicit models are normally used for predicting both the context and environment in which the alternatives are to be implemented and their associated consequences.This is necesary because the factors are usually so numerous and their interrelations so complex that intuition and mental models are not adequate to handle the large number of factors and their intricate relations.Predicting consequences is not the only,or even the first,use of models in a system analysis.It is however,the most prominent use,for such models are likely to be elaborate and programmed for a computer,whereas many other models may be no more than well-thought through concept.In our example many different models are needed to estimate the results for the alternatives,and their consequences are of different types.In our example an early problem for the analysis is to fend a way to turn the vague goal of increased highway safety into something of a more operational character——in other words，to settle on a way to measure it .Onemeasure might be the reduction in annual number of fatalities;another might be the reduction in the annual(monetary)cost of highway accidents to the victims.Another task for the analyst is to check the alternatives for feasibility.It might turn out the alternative of automated seat belts is not feasible owing to public unacceptability.If this alternative were far superior to all the others i n increasing safety,the decisionmaker would probably want to investigate the cost and effectiveness of a campaign to change public opinion.The analyst will also want to search for and examine alternatives not on the original list——such things as better emergency ambulance service,eliminating unguarded railroad crossings,and changed car design——for these alternatives may promise increased highway safety at less cost than those on the original list,and,when presented with supporting calculations,may lead the decisionmakers to expand the list of possibilities they are welling to consider.Indeed,the discovery,invention,or design of new and better alternatives is often the real payoff from systems analysis.In predicting the results associated with the various alternatives,the analyst may have to use radically differing models or methods.A model to show the effect of improved driving skills on the number of fatalities can differ considerably from a model to predict the way a lower speed limit affects fatalities.On the other hand ,predictions for both cases may be obtained statistically from experiences in other jurisdictions with similar driving conditions.In comparing alternatives various future contexts may also have to be considered,with predictions or conjectures made about the effects of,for instance,a petroleum shortage on automobile traffic and other exogenous factors beyond the decisionmakers' control.One run-through of the set of procedures is seldom enough.Several cycles or iterations are almost always necessary to refine the first models and assumptions,and thus increase one's confidence in the outcomes.系统分析的方法进行系统分析时，人们一定认为存在某个问题，或至少应该认识到存在某个问题的可能性。