Зеленская_резюме_англ (1138376), страница 2
Текст из файла (страница 2)
If an excessive number of variables isused to analyze a small number of DMUs, then Pareto efficiency frontier will be defined by abigger number of objects that will be defined as efficient DMUs. The use of principal componentshelp to reduce the number of variables while losing little information, without the need to removecomplete variables from the analysis because principal components describe key characteristics ofthe sample.Fourthly, we try to distinguish internal factors and theatre characteristics that affect thedemonstrated level of efficiency.
In order to explain the differences in efficiency levels oftheatrical institutions we use regression analysis and control for the year of observation. Weassume that not only inputs but also actions of organizations (understood in this research asoutputs) affect their efficiency.Fifthly, we try to find out statistically whether efficiency levels of theatres affect consumersatisfaction level and when this effect becomes apparent. For that purpose we use regressionanalysis and data from www.tripadvisor.com.7Suggested methodological approach is described in Figure 1.Analysis of Performance IndicatorsMethod: principal component analysisVariables: original statistical indicatorsResult: latent constructs (principal components) are identifiedAnalysis of Objects in the SampleMethod: k-means cluster analysisVariables: mean values of principal componentsResult: homogeneous groups of objects (sub-samples) are identifiedEfficiency Measurementwithin Cluster 1Method: DEAOptions: input-oriented oroutput-oriented models;returns to scale optionsVariables: principalcomponents of inputs andoutcomesResults: efficiency indicatorsof objects within Cluster 1are identifiedEfficiency Measurementwithin Cluster 2Method: DEAOptions: input-oriented oroutput-oriented models;returns to scale optionsVariables: principalcomponents of inputs andoutcomesResults: efficiency indicatorsof objects within Cluster 2are identifiedEfficiency Measurementwithin Cluster n…Method: DEAOptions: input-oriented oroutput-oriented models;returns to scale optionsVariables: principalcomponents of inputs andoutcomesResults: efficiency indicatorsof objects within Cluster n…are identifiedAnalysis of Influence ofOutput Indicators onEfficiency within Cluster 1Analysis of Influence ofOutput Indicators onEfficiency within Cluster 2Method: regression analysisVariables: efficiencyindicators of objects withinCluster 1 and originalstatistical indicators ofoutputsResults: internal factors thatinfluence efficiency areidentifiedMethod: regression analysisVariables: efficiencyindicators of objects withinCluster 2 and originalstatistical indicators ofoutputsResults: internal factors thatinfluence efficiency areidentifiedAnalysis of Influence ofOutput Indicators onEfficiency withinCluster n…Method: regression analysisVariables: efficiencyindicators of objects withinCluster n… and originalstatistical indicators ofoutputsResults: internal factors thatinfluence efficiency areidentifiedAnalysis of Impact of Efficiency Level on Consumer SatisfactionMethod: regression analysisVariables: efficiency indicators within the whole sample and ranking onwww.tripadvisor.comResults: impact of efficiency level on consumer satisfaction is identifiedFigure 1 – Suggested Methodological Approach to Efficiency Measurement of CulturalOrganizationsSource: Author’s own elaboration8Main FindingsIn order to reach the main aim and objectives of the research the author of the dissertationhas developed and empirically tested an original methodological approach to relative efficiencymeasurement of cultural organizations.
The main findings may be outlined as follows.Firstly, in the dissertation existing theoretical approaches to the notions of performanceand efficiency applicable to the cultural sphere were deeply studied. The author has shown that themajority of scholars stick to the notion of socio-cultural efficiency that is understood as a ratiobetween social, economic and cultural effects and the resources that are used to reach them.Secondly, the author has analyzed current trends that explain the necessity of performancemeasurement of cultural organizations.
The main trends are: reduction of budget funding andtighter restrictions on budget spending; need to create a multi-channel financial model of culturalorganizations; demand for clearer accountability to the funding parties; increasing competitionlevel coupled with no increase in demand for cultural services; demand for increased transparencyof organizations funded from public money and equal access to cultural merit goods.Thirdly, the dissertation contains an overview of theoretical and empirical approaches toperformance measurement of cultural organizations.
A special focus is laid on research of relativeefficiency based on econometric methods. The author shows that due to a special nature of culturalorganizations non-parametric methods are a suitable instrument for efficiency measurement.Fourthly, the author has proposed a multi-stage methodological approach to performancemeasurement of cultural organizations that allows: a) to conduct comparative analysis of a largesample of heterogeneous objects; b) to use a multi-criteria approach and quantitative indicators; c)to study different aspects of performance depending on the indicators used in analysis; d) toelaborate standardized efficiency indicators, to judge about the measures of necessaryimprovements of particular indicators and to distinguish benchmarks; e) to take a complexapproach to the problem of performance measurement – to identify indicators that influenceefficiency and to evaluate the influence of efficiency on consumer satisfaction level.Fifthly, based on descriptive analysis of performance indicators of Russian theatres,several positive and negative trends in the development of theatrical industry were identified,namely: increase in financial resources of theatres in real terms; increase in the number ofperformances; increase in attendance and revenues from ticket sales; continued dependence onbudget funding; little emphasis on sponsors and donors as a funding source; disproportion in accessto theatrical services based on income.9Sixthly, using principal component analysis the author has defined the key performanceindicators of cultural organizations that are important for relative efficiency measurement.
A setof 12 principal components has been identified.Seventhly, using cluster analysis four theatre groups (clusters) were identified (Figure 2).We described their key features from the point of view of performance indicators and real-lifecharacteristics such as status, location, genre, etc. Interpretation of these results allowed the authorto name the clusters in the following way: ‘exemplary theatres’, ‘classical academic theatres’,‘advanced regional theatres’ and ‘children theatres and theatres on the periphery’. The followingtendency was identified: the lower the number of objects in the cluster is, the higher mean valuesthose objects demonstrate. A prominent exception is the principal component ‘work with children’where an opposite tendency is observed.Figure 1 – Mean values of principal components for four clusters of Russian theatresSource: own elaborationEighthly, efficiency levels of theatres belonging to two clusters (1071 observations) werecalculated according to different DEA models depending on the outcome indicators used in theanalysis.
Comparison of mean values of efficiency indicators in 2012-2016 (Figure 3) shows thatthe values do not differ significantly. Theatres of Cluster 3 are more efficient in Model 1“Attendance” and Model 2 “Revenues”, while theatres of Cluster 2 are more efficient in Model 310“Work with Children”. These results are in tune with the findings of cluster analysis of the wholesample: “classical academic theatres” do not pay enough attention to children as target audiencecomparing with organizations from bigger clusters.0.7840.80.7810.780.7520.7440.760.7330.74Cluster 20.72Cluster 30.6920.70.680.660.64Model 1 'Attendance'Model 2 'Revenues'Model 3 'Work withChildren'Figure 3 – Mean values of efficiency indicators (Clusters 2 and 3, years 2012-2016,variable returns to scale DEA models)Source: own elaborationThe author has identified statistically significant average and strong correlation betweenthe results of efficiency measurement in different models (Tables 2 and 3).
![](https://s.studizba.com/z.php?f=/templates/si/i/noavatar.png&w=100&h=100&t=1"){kind=avatar}
