Evaluation of the Impact of Profile Direction to the Indicators of the Activities of Russian Universities

Abstract 
Since the Ministry of Education and Science of the Russian Federation began monitoring the university effectiveness, such a radical approach to cleaning up the higher education system from "useless educational institutions" has become a permanent subject of criticism in the scientific community. The main argument is that most of the monitoring indicators do not reflect the effectiveness of the institution in terms of the ratio of the output to the cost. Nevertheless, these 
indicators characterize different aspects of the activities of universities. And the results of monitoring allow to obtain statistical estimates of the degree of influence on these indicators of the specialization of the university (the structure of the student contingent in the areas of training). However, due to the specifics of the compositional data as covariates it is not possible to estimate the parameters of the regression model using the standard least squares technique. In addition, the original data are heterogeneous and contain outliers. To solve these problems, a number of robust principal component regressions are estimated. The estimation results are analyzed in terms of consistency. It is revealed that an increase in the share of mathematical, natural-science, information and engineering specialties in the structure of student training on average leads to an increase in the indicators of scientific activity, while for the economy and tourism, the opposite trend is characteristic. The larger share of the economy and tourism in the structure of specialties corresponds, on the average, to the smaller average score of the Unified State Examination and higher values of the indicator of financial and economic activity. Thus, the availability of these specialties in the structure of training is justified from the point of view of the financial status of universities. 
 
Keywords 
University, monitoring, indicator, compositional data, principal component regression, α-transformation, robust estimation 
 


References 

1. Vinokurov M. A. Monitoring the effectiveness of Russian universities: enhancing the methodology.  Izvestiya Irkutskoy gosudarstvennoy ekonomicheskoy akademii [Bulletin of Baikal State University], 2013, № 6, p. 5–11. (in Russ.) 
2. Gurtov V. A., Pitoukhin E. A., Nasadkin M. Yu. The effectiveness of higher education institutions in terms of graduates employment. Vysshee obrazovanie v Rossii [Higher Education in Russia], 2013, № 10. p. 19–27. (in Russ.) 
3. Prokhorov S. G., Svirina A. A. Efficiency of university monitoring and Russian small cities perspectives. Vysshee obrazovanie v Rossii [Higher Education in Russia], 2014, № 11, p. 121–125. (in Russ.) 
4. Dokuchaev I. I. Monitoring of the performance and quality of universities as a method of radically reforming Russia’s higher education system: political and legal aspects.  Uchenye zapiski Komsomol'skogo-na-Amure gosudarstvennogo tekhnicheskogo universiteta [Scholarly Notes of Komsomolsk-na-Amure State Technical University], 2015, № 3 (23), p. 91–95. (in Russ.) 
5. Sirotkin G. V. Elements of the new education quality control system, education and efficiency assessment of any type higher education. Tekhnicheskie nauki – ot teorii k praktike [Engineering – from theory to practice], 2013, № 26, p. 43–50. (in Russ.) 
6. Chernyshov S. Revolutionary expediency.  Ekspert-Sibir' [Expert-Siberia], 2018, April 24. 
7. Varshavskaya E. Ya. Russian employees with higher education: analysis of areas of study. Voprosy statistiki, 2016, № 9, p. 65–74. (in Russ.) 
8. Reina D. G. et al. Improving discovery phase of reactive ad hoc routing protocols using Jaccard distance. The Journal of Supercomputing, 2014, vol. 67, № 1, p. 131–152. 
9. Hampel F. R., Ronchetti E. M., Rousseeuw P. J., Stahel W. A. Robust statistics: the approach based on influence functions. John Wiley & Sons, 2011. 
10. Hron K., Filzmoser P., Thompson K. Linear regression with compositional explanatory variables. Journal of Applied Statistics, 2012, vol. 39, p. 1115–1128. 
11. Egozcue J. J., Pawlowsky-Glahn V., Mateu-Figueras G., Barcelo-Vidal C. Isometric logratio transformations for compositional data analysis. Mathematical Geology, 2003, vol. 35, № 3,  p. 279–300. 
12. Aitchison J. The statistical analysis of compositional data. Journal of the Royal Statistical Society. Series B (Methodological), 1982, vol. 44, p. 139–177. 
13. Tsagris M. Regression analysis with compositional data containing zero values. Chilean Journal of Statistics, 2015, vol. 6, № 2, p. 47–57. 
14. Timofeeva A. Yu. Robust principal component regression on compositional covariates with application to educational monitoring. Applied methods of statistical analysis.  Nonparametric methods in cybernetics and system analysis (AMSA'2017): proc. of the intern. workshop, Krasnoyarsk, 18–22 Sept. 2017. Novosibirsk, NSTU Publ., 2017, p. 241–248. 
15. Bair E., Hastie T., Paul D., Tibshirani R. Prediction by supervised principal components. Journal of the American Statistical Association, 2006, vol. 101, № 473, p. 119–137.