Page 7 - 6880

n
 (U  M i )
i
X  i 1 , (1.5)
a
n
where M – number of estimates in the range; u – mean value of interval.
i
i
After the results obtained the mandatory element is the calculation of
accuracy of expert opinion. To this end, the degree of unanimity experts is
determined by:
- double rank correlation coefficient;
- coefficient of concordance (variance or entropy);
- coefficient of variation (V), which shows the degree of deviation of the
obtained value and determined by the formula:

V  , (1.6)
X
where σ – standard deviation:

n
 (C  X ) 2
i
  i , (1.7)
n
where X – arithmetic mean of the safety indicator.
If the coefficient of variation V≤0,1 – level of experts unanimity is
considered high; 0,1<V≤0,15 – above average; 0,15<V≤0,25 – average;
0,25<V≤0,35 – below average; V>0,35 – low (in this case, the coefficient of
variation indicates a wrong selection of expert group, the high heterogeneity of the
degree of competence and result of survey is not valid).

4 TASK OF CLASS
4.1 Before meeting on class, students should study basic theory of the expert
method of risk determination. All students are expected to come to class alert and
ready to participate in discussion.
4.2 Students in the classroom at the request of a teacher give answers on the
questions.
4.3 Then the students should solve the problems and tasks (should determine
the safety level of a new product according to the experts survey) according to the
variant given by teacher. They may use references [1, 2] and theoretical data from
basic theory.
Task 1.1 Calculate the validity of expert opinion obtained in the evaluation
of two sample of new equipment. Determine which of the two products have a
greater level of safety. Compare auxiliary mean value of the indicator and mean-
arithmetical value. The results of the expert survey are shown in Table 1.1

Table 1.1 - Data for solving Task 1.1

Expe rts 0 1 2 3 4 Variant 6 7 8 9 10
5

7