1st ICAI 2020

International Conference on Automotive Industry 2020

Mladá Boleslav, Czech Republic

Previously, the variables of the first group (structural variables) were selected from the database. Then, using the IBM SPSS program, a matrix of paired Pearson mutual conjugation coefficients was constructed, with the help of which a second group of variables (variables characterizing the results of economic activity) was selected based on the conditions for ensuring significant correlation with structural variables. The calculation of the matrix of paired Pearson mutual conjugation coefficients is presented in Table 3. Table 3: The matrix of pairwise Pearson mutual conjugation coefficients X 1 X 2 X 3 X 4 X 5 X 6 X 7 X 8 X 1 1 0,357** -0,081 0,169* 0,144* 0,127 0,248** 0,308** X 2 0.357** 1 0,725** 0,767** 0,731** 0,700** 0,895** 0,964** X 3 -0.081 0,725** 1 0,699** 0,680** 0,651** 0,684** 0,672** X 4 0.169* 0,767** 0,699** 1 0,889** 0,860** 0,905** 0,798** X 5 0.144* 0,731** 0,680** 0,889** 1 0,980** 0,835** 0,774** X 6 0.127 0,700** 0,651** 0,860** 0,980** 1 0,806** 0,745** X 7 0.248** 0,895** 0,684** 0,905** 0,835** 0,806** 1 0,943** X 8 0.308** 0,964** 0,672** 0,798** 0,774** 0,745** 0,943** 1 ** The correlation is significant at the level of 0.01 (2 sides). Source: own calculations Cluster analysis was carried out using the Word method, which allows to split the whole volume into a sufficient number of clusters corresponding to the economic nature of the studied phenomena. Due to different units of measurement, the studied indicators were preliminarily standardized. Objects with missing data were excluded from the analysis, as well as objects with abnormally high figures of added value and total assets in comparison with other enterprises. So, cluster analysis was carried out on a sample of 222 enterprises manufacturing electrical equipment, electronic equipment, as well as other products and equipment for motor vehicles. The hypothesis of equality of variances within and between clusters is rejected for all variables at 5 and 216 degrees of freedom. The value of p (the probability of error in accepting the hypothesis of inequality of variances) is not more than 0.01 (the F-criterion is significant for all variables at a level of at least 0.01). This fact suggests that the hypothesis of dispersion inequality is accepted and, accordingly, clusters are formed correctly. The result of the cluster analysis was the division of 222 enterprises for the production of electrical equipment, electronic equipment, as well as other products and equipment for motor vehicles into 6 clusters. The results of the cluster analysis of enterprises for the production of vehicles are presented in table 4.

34