Search and selection of experts

The composition of experts involved in forecasting is a key parameter determining the quality of the future forecast, for this selection of experts it is necessary to pay close attention.

It is possible to identify the initial circle of persons who are potential candidates for experts by publications, since authors writing about the object of forecasting, as a rule, are not only well acquainted with the object itself, but are often the most competent people on this issue. The second way to identify candidates for experts is to survey potential candidates among specialists who are even remotely familiar with the object of forecasting. If the list of candidates obtained at this step does not satisfy the organizers of the survey, then for its subsequent expansion it is necessary to ask a question about the most suitable candidates to all those selected in the previous stage. If the importance of the forecast is large enough, then this question should be asked to all newly named candidates. The selection process can be considered fully completed, once the newly named candidates name mostly those already on our list.

After the formation of the list of candidates from it is completed, it is necessary to select experts. The simplest method of selection is to select those candidates whose names were most frequently mentioned at the stage of searching for candidates for experts.

The second method is the self-assessment procedure or, much better, the evaluation by each candidate of all other candidates. When experts evaluate each other, the results of the assessment are much more objective than with self-assessment, but in this case the candidates should know each other well. If this condition is not met, a self-assessment procedure is applied. At the same time, those participating in the evaluation or self-assessment should be offered criteria against which candidates will be evaluated and it is necessary to indicate the scale against which the assessment should be made. The criteria can be the degree of awareness of the expert with theoretical issues related to the object of forecasting, the experience of practical work in this area, the degree of development of intuition for predicting the behavior of the forecasting object, the depth of acquaintance with literary sources. To get a final score, each of the criteria must be assigned a normalized weight. Weights are considered normalized if their sum is equal to one. The specific weight values for each criterion can be asked to everyone involved in the assessment or self-assessment. The sum of the works obtained for all the evaluation criteria for each examiner on the weights of these criteria is the final assessment of the expert.

Finally, the last method of selecting experts from the roster of candidates is the method of security questions. In this case, candidates are asked questions about the forecasting object, the answers to which are not known to the candidates, but are known exactly to the organizers of the forecast. Usually these are questions about not very important, private characteristics of the forecasting object in the past, which, due to their insignificance, are not remembered by most specialists, but the organizers of the survey know them exactly from any documents (directories, reports, etc.). For example, when selecting experts to predict the number of the armed forces of the Republic of Belarus for the future, the question of the number of women servicemen in the armed forces in 1995 can serve as a control issue, when forecasting the development of mobile communications – the number of phones in rural areas in such and such a year, etc. At the same time, if there is a suspicion that the figure is sufficiently known, there is always the possibility of its further detailing, for example, ask a question about the number of phones in rural pensioners and the like. The best experts are those candidates who have given the most accurate answers to security questions.

One of the most important problems in the selection of experts is the task of the minimum required number of them. Unfortunately, the theory does not give a clear answer to this question. An answer is only possible about whether there are enough experts, but this answer can be obtained only after processing the results of the survey. The minimum number of examiners can be estimated by the formula:

where:         – the minimum required number of experts;

        – Relative change in the average value of the experts’ answers when adding another expert;

        – the average value of experts’ answers;

        – the average value of the experts’ answers when adding another expert;

               – the maximum value in the answers of experts.