Modeling of publication performance using a system dynamics approach: Evidence from Kazakhstan

  • 18 Views
  • 1 Downloads

Creative Commons License
This work is licensed under a Creative Commons Attribution 4.0 International License

Type of the article: Research Article

Scientific productivity is one of the key benchmarks to evaluate the development of a country in terms of innovation potential and competitiveness in the global academic environment. This study investigates the impact of key factors such as faculty workload and satisfaction, career dynamics, institutional environment, and research funding on publication performance. The methodology is based on system dynamics (SD), a simulation approach that assesses the dynamic impact of multiple factors on the publication performance. The approach combines causal-loop analysis, stock-and-flow diagrams, sensitivity testing, and Monte Carlo simulations, with Kazakhstan serving as a case country. The data on the factors, including from the global literature and statistics from Kazakhstan, were simulated to forecast publication performance and research output through 2030. Sensitivity analysis identified six dominant factors affecting the number of researchers: average wage growth rate, Ph.D. growth rate, average author tenure, normal workload, job satisfaction, and productivity rate. In contrast, the number of publications was most influenced by the research expenditure per paper, government investment growth rate, and private sector investment growth rate. The study findings from Kazakhstan can serve as “lessons learned” and assist researchers and policymakers working in the science policy and evaluation area in identifying high-impact factors and formulating more effective strategies to enhance the scientific knowledge creation of a country.

Acknowledgments
This paper has been funded by the Science Committee MSHE RK (Grant “Modernization of the quality assurance system of higher education in Kazakhstan based on digitalization: development of approaches, mechanisms and information base” BR24992974).

view full abstract hide full abstract
    • Figure 1. Factors affecting publication performance (SD modeling created)
    • Figure 2. Sensitivity analysis of the subsystem (Authors)
    • Figure 3. Sensitivity analysis of the subsystem (Papers published)
    • Figure 4. MC simulation results for the number of authors under ±10% variation
    • Figure 5. MC simulation results for the number of authors under ±20% variation
    • Figure 6. MC sensitivity graphs of the stock variable Papers published under ±10% variation
    • Figure 7. MC sensitivity graphs of the stock variable Papers published under ±20% variation
    • Table 1. Variables in the subsystem (Authors)
    • Table 2. Variables in the subsystem (Papers published)
    • Table 3. Control limits for the MC simulation of the Authors and Papers published
    • Table 4. MC forecast of the number of authors in 2030
    • Table 5. MC forecast of the Papers published in 2030
    • Table A1. Critical factors used for SD modeling
    • Formal Analysis
      Andrey Samoilov, Kuralay Nurgaliyeva
    • Investigation
      Andrey Samoilov, Anel Kireyeva
    • Methodology
      Andrey Samoilov, Kuralay Nurgaliyeva, Anel Kireyeva
    • Software
      Andrey Samoilov, Anel Kireyeva
    • Validation
      Andrey Samoilov
    • Conceptualization
      Kuralay Nurgaliyeva, Yerzhan Mukashev, Anel Kireyeva
    • Project administration
      Kuralay Nurgaliyeva, Yerzhan Mukashev
    • Resources
      Yerzhan Mukashev
    • Supervision
      Yerzhan Mukashev, Anel Kireyeva
    • Writing – review & editing
      Yerzhan Mukashev
    • Data curation
      Anel Kireyeva