Page 12 - Azerbaijan State University of Economics
P. 12

THE JOURNAL OF ECONOMIC SCIENCES: THEORY AND PRACTICE, V.78, # 2, 2021, pp. 4-16



                    The main disadvantage of the moment coefficient of asymmetry is that its value is
                    very  sensitive  to  the  sharp  difference  of  any  certain  value  of  the  variable  in  the
                    collection. The structural coefficient of asymmetry characterizes the asymmetry in
                    the central part. Unlike the moment coefficient of asymmetry, any sharp difference
                    in the value of variables does not affect the value of this coefficient. In practice, the
                    structural coefficient of asymmetry proposed by Karl Pearson is widely used.

                                                             x  −  mod
                                                    As  pirson  =  


                    In a variation sequence, if mode=median=mean, this means the sequence is symmetric.

                    If the mode of a variation sequence is less than mean, then the structural coefficient
                    of asymmetry becomes greater than 0, and thus the sequence is asymmetric to the
                    right. By contrast, if the mode of a sequence is grater than mean, in this case, the
                    structural coefficient of asymmetry becomes less than 0, which means the sequence
                    is asymmetric to the left.

                    RESULTS
                    Testing distributions of scores
                    For  example,  we  will  test  the  admission  scores  of  bachelors  for  the  2017/2018
                    academic  year  on  the  basis  of  table  3,  by  Pearson's   criterion,  to  find  out  if  the
                    admission scores are distributed normally. For this we will make the table 4.

                           Table 4: Testing normality of scores obtained by bachelors in the year of 2017
                                                observed                         expected     (O-
                        no    interval   up to b  frequnecy   p(x<a)   p(a<x<b)   frequency   E)^2/E
                        1      0-10      10       22       0.00551    0.00551   116.14877   76.31584
                        2      11-20     20       421     0.031548   0.026038   548.82669   29.77199
                        3      21-30     30      2244     0.120001   0.088452   1864.4009   77.28782
                        4      31-40     40      4628     0.311552   0.191551    4037.517   86.35758
                        5      41-50     50      5415     0.576157   0.264605   5577.3384   4.725149
                        6      51-60     60      4361     0.809376   0.233219   4915.7896   62.61284
                        7      61-70     70      2540     0.940517   0.131141   2764.1878   18.18261
                        8      71-80     80      1174      0.98754   0.047024    991.1668   33.72589
                        9      81-90     90       246     0.998284   0.010744   226.45914   1.686155
                        10    91-100     100      27      0.999847   0.001562   32.934262   1.069265
                                                  21078                                    391.7351
                      mean       47.19
                      st.dev     14.63

                                                 Source: authors’ calculations


                                                           12
   7   8   9   10   11   12   13   14   15   16   17