Task participants

The interrogation of experts as a certain stage of their work is often perceived by DMP as a certain "zero" cycle of an investigation.

In conclusion of this review it is necessary to mention that the shortcomings of some subjects of the expert prognostication are engendering obstacles for the others and, in the end, are embodied in mistakes including systematic ones.

lecture № 2.SYSTem AnaLYSIS OF EXPERT PROGNOSTIC TECHNOLOGY

Prognoses are being made either directly on the basis of experts’ opinions or indirectly - on the basis of a model of an object or process being forecasted developed with the help of experts.

In this lecture we shall discuss an expert prognostication. It takes diverse forms and may play a different role in various socio-economic conditions.

As an activity expert prognostication depends sufficiently on the social infrastructure within the framework on which the decisions are made or realized. Under some conditions it may be just a formal and demonstrative act while degenerating into an imitation of a genuine prognostication, and under other ones, just on the contrary, it serves as an efficient factor within the management system.

The role of prognostication depends both on the character of the management itself and on the type of responsibility for decisions making, and on the mentality dominating in the society and on the level of the progress of democracy.

A forming of scientific premises for making and practically realizing management decisions on the basis of this prognostication under the evolution uncertainty conditions of the object or process being forecasted, we shall consider as the purpose of the expert prognosis elaboration. Other purposes as, for example, an influence upon the object of the prognosis through a wide proliferation and specific usage of perspective appraisals (an active prognosis) we purposefully leave outside the frameworks of our discussion.

In connection with the fact that at present the price of the quality and, correspondingly, error of a prognosis is rising more and more, we are addressing the problems of expert prognostic technology. It is impossible to assert that there are no similar attempts in literature. It is true that they are being realized in the sphere similar to the prognostication that is to the examination. For example, in the work of S.D. Beshelev and F.G. Gourvich there were singled out the following "basic stages of conducting an examination a succession and specific contents of which is changing depending on the purposes of the examination:

· a formulation of the purpose of the examination and an elaboration of an interrogatory procedure;

· a forming of a group of specialists-analysts (organizers of the expertise);

· a selection and formation of the group of experts;

· a realization of the interrogatory procedure;

· an analysis and processing of information obtained, from experts;

· a synthesis of an objective a priori information and information obtained as a result of the expertise with the purpose of putting them into a form convenient for decision making." [9].

We shall accept the scheme of S.B. Beshelev and F.G. Gourvich as a good basis for discussing and try to develop it. Certainly, a great variety of problems in working out even simple prognoses hampers a formalization of the procedure of their preparation and realization. A realization of large-scale complex prognoses is becoming complicated over and over. The problems engendered in this case are discussed in the source [20].

It is quite natural that the entire process of expert prognostication should be accurately organized and should have a definite technology which is discussed below.

In the modern practice the person who orders a prognosis or (and) DMP may make decisions on the basis of his own judgements. In other situations DMP, while not being a unique specialist on the problem and being not sufficiently competent in it generally (and while understanding this fact), may enroll experts "directly". At the same time it is becoming more and more evident that the presence of subjects of two types -"Expert" and also "DMP" or (and) a customer of the prognosis (further in the text - "Customer"- and they are often indivisible) - it is not enough for solving numerous prognostic problems.

In addition to experts it is also necessary to involve prognostications development organizers (further in the text - "Organizer") on a permanent basis and also specialists in the sphere of a methodological and methodic support of the prognostication ("Prognosist") who are playing the role of technologists as a matter of fact.

An expansion of the number of subjects participating in the prognostication and also an approach towards expert prognostication technology as an object of the system analysis allows to raise the quality of prognoses.

Certainly, the system approach towards the object of prognostication is also necessary.

And this system character should become apparent in setting limits within the frameworks of which the object would be considered, in a clear cut setting of expert prognostication aims and goals, in taking external relations of the object being studied into account.

The system character should be also expressed in the fact that expert prognostication, as a kind of activity, is an element of the general process of substantiation and making decisions.

We consider, the main point is that it is necessary to represent expert prognostic technology as a system, that is as a certain unity of elements tied up into the whole with some relations (in a particular case - with interactions).

In addition, it is necessary to reveal system-forming elements and opportunities of their transformation into system-making ones. Therefore, the system character would be realized in a two-step selection of elements of the technology of getting an expert conclusion.

In the first step of the development of a prognosis, it is necessary to single out some sub-multitudes from several multitudes of elements of the technology considered as a system and possessing "system-significant" properties.

In the second step the elements possessing "system-defined" properties are being picked out from the above mentioned ones.

A selection of elements is realized while taking the following into account (three conditions of choice):

1. A specificity of the object or process under consideration;

2. Limitations of various kinds:

· financial ones (payment for the entire work, for independent experts);

· personnel ones (possibilities of selecting experts, organizers of prognostication, prognosists);

· temporal ones (the period required for an organization and realization of the prognosis);

3. Level of the problem understanding by DMP and the goals of Customer and DMP.

From our point of view the basic system-forming elements of expert prognostic technology are as follows:

1. Methods of forming an expert group. (In the works [79,100] such methods are indicated as mutual recommendations, snow ball and others).

2. Methods of interrogating the experts (in particular, filling in questionnaires, interviewing);

3. The simplest types of expert appraisals (including verbal, mark, interval, numerical ones [79]);

4. Methods of selecting the most skilled experts into a group and the rules of withdrawing them from its staff [80];

5. Algorithmic operations which allow to obtain a chosen type of the expert appraisal (for example, the Churchman-Ackoff or fon-Newman - Morgenstern methods described in the monograph [49]);

6. The simplest methods and procedures which allow to obtain the knowledge from experts (for example, the brain attack or "Delfi" method);

7. Forms of the error index for an a posteriori appraisal of the expert conclusion quality (for example, such ones as E(x,y)= |x-y|, where x is an expert appraisal; y is a true value).

8. The basic principles of carrying out expertise (for example such ones as independence of experts on other subjects of expertology).

While using a new approach let's consider a general technology of working up the expert prognosis. A realization of each of the stages of this diagram and procedures within one or another stage is accomplished by the above-mentioned subjects. A part of these procedures, especially organizational ones, should be regulated. This thesis concerns the procedures of the expertise ordering, selecting the participants of prognostication, interrogating the experts and making a good use of expert conclusions.

At every stage and in any procedure the subject is allotted its share of rights and responsibilities depending on its functional role in the process of getting a high-quality expert conclusion.

The 12-stage diagram suggested by us differs from the above-mentioned list of the expertise realization stages not only with the composition and detail of the description, but also with the indication of relations between stages and the connection of separate stages with the object of prognostication. It allows to reflect a continuity of the prognostication elaboration process in a better way which is also important for getting high-quality prognoses.

Certainly, while having chosen just these twelve stages of expert prognosis development we should have in mind that a succession and specific contents of every stage may and must be changed depending on the conditions of choice.

Let's consider the diagram of working up expert prognostic technology (Pict.1).

Incoming

information