CC BY
0
CLUSTERING OF KAZAKHSTAN^ UNIVERSITIES BASED ON QUALITY INDICATORS
Yelmar Orazbek, Master of Economics, Senior Lecturer1, PhD student2 international IT University 2"Turan" university (Kazakhstan, Almaty)
DOI:10.24412/2500-1000-2024-6-3-36-40
Abstract. This study presents a clustering analysis of universities in Kazakhstan based on various quality indicators, including academic reputation, employer reputation, faculty-student ratio, citations per faculty, and internationalization metrics. The aim is to identify groups of universities with similar characteristics to facilitate targeted improvements and strategic planning. The Ward's method of hierarchical clustering was employed, resulting in four main clusters. The findings reveal distinct strengths and weaknesses of each cluster, providing valuable insights for policymakers and university administrators to enhance the quality of higher education in Kazakhstan. The study concludes with specific recommendations for improving academic reputation, international cooperation, infrastructure, and employer engagement.
Keywords: clustering, universities in Kazakhstan, academic reputation, quality of education, international research, Ward's method, higher education analysis, university performance indicators, educational policy, internationalization.
In today's world, the quality of higher education is vital for the development of society and the economy. Universities play a crucial role in advancing knowledge, fostering innovation, and preparing skilled professionals [1]. Thus, evaluating and enhancing the quality of universities is of utmost importance. In this context, clustering universities based on various performance indicators offers valuable insights into their strengths and weaknesses, aiding in targeted improvements and strategic planning [2].
Kazakhstan, with its rapidly evolving educational sector, has made significant investments in its higher education institutions. However, the varied quality and performance of these institutions call for a detailed analysis to identify leading universities, those with balanced performance, and those facing challenges in specific areas. This study aims to cluster universities in Kazakhstan using a comprehensive set of quality indicators, including academic reputation, employer reputation, faculty-student ratio, citations per faculty, and internationalization metrics, among others [3].
The methodology used for clustering employs Ward's method, a hierarchical clustering technique known for its effectiveness in
creating clusters with minimal variance. This approach ensures that the resulting clusters are both statistically robust and practically meaningful, providing a clear basis for policy recommendations and strategic planning [4].
By identifying clusters of universities with similar characteristics, this study aims to offer actionable insights that can help policymakers, university administrators, and other stakeholders enhance the quality of education and research in Kazakhstan [5]. The findings of this study are expected to contribute to the ongoing efforts to elevate Kazakhstan's higher education institutions to international standards and improve their competitiveness on the global stage [6].
The following sections detail the study's methodology, present the results including the identified clusters and their characteristics, and provide an in-depth analysis of the clusters with recommendations for improving the quality of education in Kazakhstan's universities.
Methodology
The study aimed to cluster universities in Kazakhstan based on various quality indicators to identify groups with similar characteristics. The following methodology was employed:
- Экономинеские uayHU -
Data Collection
The data for this study was collected from multiple sources, including university reports,
academic databases, and international rankings. The indicators used for clustering included (Figure 1).
Academic Reputation Employer Reputation Faculty-Student Ratio Citations per Faculty
International Faculty International Students International Research Network Employment Outcomes
Sustainability Overall Performance Teaching Quality Research Environment
Research Quality
Industry Collaboration
International Outlook
Fig. 1. The indicators used for cluster analysis
These indicators offer a comprehensive perspective on the quality and performance of each university [7].
Data Processing
The collected data underwent processing and normalization to ensure consistency across different indicators. This normalization was crucial to account for the varying scales and units of the indicators. Each indicator was converted into a standardized z-score.
Clustering Method
Ward's method of hierarchical clustering was selected for this analysis. Ward's method is a widely used clustering technique that minimizes variance within each cluster [8]. It achieves this by merging the pair of clusters that leads to the smallest increase in the total within-cluster variance at each step. This
method is particularly effective for creating clusters of roughly equal size and minimizing the sum of squared differences within all clusters.
Cluster Validation
The resulting dendrogram (cluster diagram) was analyzed to determine the optimal number of clusters. The optimal number of clusters was identified based on the point where the increase in within-cluster variance becomes significant, indicating that further merging would lead to overly heterogeneous clusters.
The provided cluster diagram visually represents the hierarchical clustering of universities in Kazakhstan based on various quality indicators. The diagram is divided into four main clusters (Figure 2).
Fig. 2. Clustering of the universities by Ward method
Cluster 1 (C1): This cluster stands out prominently at the top, indicating a significant level of distinction from other clusters. Universities like Al-Farabi Kazakh National University in this group demonstrate exceptional performance across various indicators, especially in academic and employer reputation (Figure 2).
Cluster 2 (C2): Positioned adjacent to Cluster 1, this group includes universities with balanced but not outstanding performance. Institutions such as Satbayev University and NSC KIMEP University are part of this cluster, displaying consistent yet average scores across most indicators.
Cluster 3 (C3): This cluster comprises universities with significant participation from international faculty and students. Universities like S.D. Asfendiyarov Kazakh National Medical University are notable for their efforts in internationalization.
Cluster 4 (C4): Situated at the bottom of the diagram, this cluster consists of universities with lower performance across most metrics. Institutions such as Kazakh National Agrarian University (KazNAU) fall into this category, facing substantial challenges and having considerable room for improvement.
The hierarchical clustering provides a clear visualization of how universities in Kazakh-
stan are grouped based on their quality metrics. This analysis highlights the strengths and weaknesses of each cluster, offering a foundation for targeted improvements and policy recommendations.
Based on the analysis conducted, four primary clusters of universities were identified. Each cluster shares specific characteristics, allowing for insights into their respective strengths and weaknesses. Below is a detailed description of each cluster.
Cluster 1 (C1)
This cluster includes top universities in Kazakhstan, such as Al-Farabi Kazakh National University and L.N. Gumilyov Eurasian National University (ENU). These institutions boast high academic reputation scores (51.4 and 29.5 respectively) and employer reputation scores (74.9 and 51.7 respectively), indicating strong teaching quality and high demand for graduates in the job market. Additionally, they excel in international cooperation, with significant percentages of international faculty (28.8% and 59.3%) and students (34.0% and 5.7%). They also perform well in metrics like the International Research Network and Sustainability, highlighting their active participation in global research and commitment to sustainable development. Overall, universities in Cluster 1 lead in both
educational and research activities in Kazakhstan.
Cluster 2 (C2)
Universities in this cluster show average performance across most metrics. Institutions such as Satbayev University and NSC KIMEP University are included here. They have balanced academic reputation scores (23.3 and 20.0 respectively) and employer reputation scores (36.0 and 30.0 respectively). Their faculty-student ratios and proportions of international faculty and students are also acceptable. However, to improve their standing, these universities need to focus on enhancing research quality and increasing citations per faculty member. Greater involvement in international projects and stronger scientific activities can elevate their academic status and competitiveness globally.
Cluster 3 (C3)
This cluster comprises universities with high proportions of international faculty and students. Institutions like S.D. Asfendiyarov Kazakh National Medical University and Auezov South Kazakhstan University (SKU) show significant international engagement, enriching their cultural and academic environment. The high presence of international faculty and students promotes intercultural communication and experience exchange, benefiting the educational process. Despite this, these universities need to improve their academic and employer reputation scores and increase the number of scientific publications and citations. This will help them strengthen their positions and attract more students and faculty from both Kazakhstan and abroad.
Cluster 4 (C4)
Universities in this cluster exhibit low performance across most metrics. Institutions
References
1. Elbawab, Roba. 2022. University Rankings and Goals: ACluster Analysis. Economies 10: 209. https://doi.org/10.3390/economies 10090209.2022.
2. Stoupas G., Sidiropoulos A., Katsaros D., Manolopoulos Y. (2021). Ranking Universities via Clustering. International Journal of Intelligent Systems Studies, 12(2), 1-15. DOI: 10.15587/1729-4061.2021.265860.
3. Kazakhstan University Management Level Clustering (2022). Cluster Analysis of the Effectiveness of University Management. Control Processes, 27, 1-15. DOI: 10.15587/17294061.2022.265860.
4. Iscte-IUL Repository (2022). University Rankings and Goals: A Cluster Analysis. Reposi-tório do Iscte, 1-15. DOI: 10.15587/1729-4061.2022.265860.
such as Kazakh National Agrarian University (KazNAU) and Suleyman Demirel University, Kazakhstan, face numerous challenges, including low academic reputation, weak employer reputation, and minimal participation in international research. To improve their standing, these universities need to focus on enhancing education quality, increasing the proportion of international faculty and students, and actively participating in international research projects. These efforts will help attract more resources, improve teaching quality, and enhance their overall reputation.
Recommendations for Improving Education Quality in Kazakhstan's Universities:
Enhancing Academic Reputation: Implement professional development programs for faculty, encourage scientific activities, and promote publications in international journals.
Improving Conditions for International Cooperation: Create conditions to attract international faculty and students, and participate in international educational and research projects.
Developing Infrastructure and Educational Programs: Update educational programs to meet international standards and improve the material and technical base of universities.
Strengthening Ties with Employers: Develop joint programs with businesses and industries, and organize internships and practical training for students in leading companies.
These measures will help Kazakhstan's universities improve their positions in international rankings and enhance the quality of education, contributing to the development of science and the country's economy.
5. Economics and Sociology (2022). Measuring the Performance of Universities through Cluster Analysis. Economics and Sociology, 1-15. DOI: 10.15587/1729-4061.2022.265860.
6. Saxena G., Lai K.A., Allen P.J. (2023). Clustering of UK universities based on the research productivity of psychology departments. Power and Education, 15(3), 385-393. https://doi.org/10.1177/17577438221130983.
7. Chen, Shun-Hsing & Wang, Hui-Hua & Yang, King-Jang. (2009). Establishment and application of performance measure indicators for universities. The TQM Journal. 21. 220-235. 10.1108/17542730910953004.
8. Matlis G., Dimokas N., Karvelis P. Unveiling University Groupings: A Clustering Analysis for Academic Rankings. Data 2024, 9, 67. https://doi.org/10.3390/data9050067.
КЛАСТЕРИЗАЦИЯ УНИВЕРСИТЕТОВ КАЗАХСТАНА НА ОСНОВЕ
ПОКАЗАТЕЛЕЙ КАЧЕСТВА
Елмар Оразбек, магистр экономики, старший преподаватель1, докторант PhD2 1Международный университет информационных технологий ^Университет «Туран» (Казахстан, г. Алматы)
Аннотация. В данном исследовании представлена кластерный анализ университетов Казахстана на основе различных показателей качества, включая академическую репутацию, репутацию у работодателей, соотношение преподавателей и студентов, цитируе-мость на одного преподавателя и показатели интернационализации. Цель заключается в выявлении групп университетов с аналогичными характеристиками для содействия целенаправленным улучшениям и стратегическому планированию. Метод иерархической кластеризации по Уорду был применен, в результате чего были выделены четыре основных кластера. Полученные результаты выявляют явные сильные и слабые стороны каждого кластера, предоставляя ценные данные для разработчиков политик и администраторов университетов с целью повышения качества высшего образования в Казахстане. Исследование завершается конкретными рекомендациями по улучшению академической репутации, международного сотрудничества, инфраструктуры и взаимодействия с работодателями.
Ключевые слова: кластеризация, университеты Казахстана, академическая репутация, качество образования, международные исследования, метод Уорда, анализ высшего образования, показатели эффективности университетов, образовательная политика, интернационализация.
