| Abstract [eng] |
Due to the different historical and geographical environment, economic and political conditions, the state of European countries’ education systems is uneven. Comparing the effectiveness of countries’ education systems is a difficult multi-criteria optimization task. This study proposes the methodology for effectiveness estimation of 29 European countries‘ education systems by applying the Composite Indicator (CI) paradigm. Given that the results of SAW (Simple Additive Weighting) aggregations can be replicated by DEA (Data Envelopment Analysis), the CIs should be constructed using the DEA model. In the context of the CIs construction, we expect a high discrimination power to enable the construction of rankings, but the DEA model showed limited discrimination power in this study. Two alternative ways to enhance discrimination power of the DEA model are analyzed – the use of weight restrictions in the DEA model and the reduction of data dimension before the DEA model. The main contribution of this study is the development of a new type of weight restriction that allows gradual transition from fixed to flexible weighting systems and solves the problem of insufficient discrimination of the DEA model. The use of hybrid PCA-DEA approach for effectiveness assessment has been investigated in this study. Despite the fact that this model has a higher discrimination power than the unrestricted DEA model when assessing the European countries' education systems, it has remained insufficient to perform ranking of the countries analysed. According to the research, the methodology for effectiveness estimation of education systems was proposed. |
