1st ICAI 2020

International Conference on Automotive Industry 2020

Mladá Boleslav, Czech Republic

In order to identify the factors affecting the size and scope of the enterprise (“number of divisions”), the following indicators were initially selected: principal business activity (NACE); revenue per place of business (CZK million); number of employees; average number of employees per point of sale (people); total assets; claims against associates; debts to associates; value added; personnel costs; depreciation; production and amortization. At construction of a tree of decisions the target variable quantity of divisions was allocated) (we shall note, that the enterprises of the analyzed cluster the quantity of divisions changes from 1 to 7, average value – 1,6). The model includes the following variables: value added; average number of employees in one subdivision (person); revenue per division (million CZK). The differences between the groups of enterprises were determined on the basis of Chi-square criterion at the level of significance not lower than 0.028. After the analysis, a model was selected which was built using the usual CHAID method. Among all possible variants of construction of decision trees the most precisely describing tree has been chosen. The decision tree with 7 nodes, including 4 terminal levels in model 2, was obtained. The root node, which consisted of 270 enterprises, contains 167 enterprises with only one unit, 61 enterprises with 2 units, 28 enterprises with 3 units and the rest (14) enterprises include 4 or more units. It may be noted that although this model is the best one constructed, the risk assessment is quite significant and amounts to 0.381, which suggests that 38.1% of observations were incorrectly classified (61.9% were correctly classified). In regard to the analysis of the scheme of the built model and terminal nodes in particular, it has already been mentioned, the enterprises are divided into 4 groups in terms of the number of subdivisions. First of all, the number of subdivisions is influenced by the value-added index. Enterprises with value added less than 42,969 thousand crowns have an average number of units of 1.4. For enterprises with value added above EEK 42,969 thousand, the average number of units was 1.9. For larger enterprises with higher value added and the number of units, the amount of revenue per unit affects the number of units. Smaller companies have an average number of employees in 1 subdivision. 4. Conclusion On the whole, analyzing the obtained results, one can draw a conclusion about their coincidence with the theoretical model of development of the organizational structure of industrial enterprises proposed by the authors. Taking into account the results of the research on the distribution of variables in clusters 1 and 4, the comparison of the results of the analysis with the theoretical model of the development of the organizational structure of industrial enterprises can be presented as follows (see Table 6).

27