|
Y
|
Y x
|
|
|
Y-Yx
|
(Y-Yx)2
|
4.2
|
3.93
|
-0.9
|
0.81
|
+0.27
|
0.073
|
4.8
|
4,90
|
-0.3
|
0.09
|
-0,10
|
0.010
|
5.3
|
5.55
|
+0.2
|
0.04
|
-0.25
|
0.062
|
6.0
|
5.95
|
+0.9
|
0.81
|
+0.05
|
0.003
|
6.2
|
6.05
|
+1,1
|
1.21
|
+0.15
|
0.022
|
5.8
|
5,90
|
+0.7
|
0.49
|
-0,10
|
0.010
|
5.3
|
5,43
|
+0.2
|
0.04
|
-0,13
|
0.017
|
4.4
|
4.78
|
-0.7
|
0.49
|
-0.38
|
0.144
|
4.0
|
3,70
|
-1.1
|
1.21
|
+0.30
|
0.090
|
|
46.0
|
-
|
5,19
|
-
|
0.431
|
Using the above formula, we define the correlation attitude characterizing the relationship between the age and the productivity of the worker:
In conclusion, it is worth noting that only two examples of double correlation were used. However, this method can be used to study the relationship between different economic indicators. This allows you to have a deeper knowledge of the situation being studied, the role and significance of each factor in its transformation.
The economic conditions and processes of the business are dependent on a number of factors. The interaction of such complex factors characterizes the situation in which the study is being conducted.
The most important is to select factors for correlation analysis . Because the final results of the analysis depend on the correct choice of factors . The practical experience of theory and analysis plays an important role in selecting factors. Here is what follows:
1. First of all, it is necessary to take into account the relationships between indicators, as they only explain the essence of the case being studied. An analysis of such factors in mathematical relations with the resultant indicator is of little consequence.
2. Creating a multifactorial correlation model requires selecting the most important factors affecting the final set of indicators, including those that can not be implemented in practice. It is not advisable to take into account the factors that are less accurate in terms of the table, the criterion of reliability.
3. It is not recommended to include factors that are related to linear characteristic linear character correlation models.
4. It is not possible to add factors that are interconnected to the correlation model. If the correlation coefficient between the two factors is greater than 0.85, one should be excluded according to the correlation analysis rule, otherwise it will result in the final indicator.
5. It is not recommended to include the correlation model with the correlation model of the functional character of the connective factors.
Comparison of analytical grouping, parallel and dynamic rows in the selection factor for correlation model, line graphs greatly helps. They can determine the presence, shape, and orientation of the interactions between the indicators studied by them. Factor selection can also be done in the process of correlation analysis, based on the value of the Stuudent criterion.
Taking into account the above mentioned conditions and using factor sampling methods, the following factors have a major impact on the level of profitability of the multi-factor correlation model ( Y ).
x 1 - flour, sums;
x 2 - savings flour, UZS;
x 3 - productivity (average annual production of a worker), thousand UZS;
x 4 - Duration of rotation of the working capital of the enterprise, days;
x 5 - Specific weight of high-quality product,%
Since the correlation link is fully visible in a large number of observations, the selection of data should be sufficiently large, since the majority of observations have the effect of other factors. The more the set of studied objects is, the more precise the results are.
Taking into consideration this requirement, the effectiveness of the factors listed above will be examined by 40 enterprises.
The accuracy, consistency and accuracy of initial data collected from each factor and the resulting set of indicators should be verified.
First of all, make sure that the information is accurate, consistent with its objective existence . The use of inaccurate materials leads to uncertainty and inaccurate conclusions.
The data should be of one level in terms of average distribution . If groups of objects differ greatly from the average, this indicates that the initial data is not of one type.
The criterion for the uniformity of the data is the average square deviation and variation coefficient calculated on each factor and output.
The average square deviation indicates absolute deviation of the mean values from the arithmetic mean values. Its value is found in this formula
The variation coefficient indicates the relative deviation of the individual values from the arithmetic mean values. It is calculated according to the following formula
The higher the coefficient of variation, the greater the scattering and the less correction of the studied objects. The variation of variables is as follows: if not exceeding 10% - insignificant, 10-20% - average, more than 20% - but not more than 33%. When the variation exceeds 33%, it indicates that this information is not of one type, and should be excluded from the non-one. These are usually at the beginning or end of the selection row .
Do'stlaringiz bilan baham: | 
