| Abstract [eng] |
Forecasting student flows is receiving increasing attention from international researchers, however in Lithuania the topic remains underdeveloped. There is a lack of studies applying statistical models to forecast student admission, even though this issue has become particularly relevant in a changing demographic and educational environment, where after a steady decline until 2022, the number of students at Lithuanian universities has begun to show an upward trend. Accurate forecasts enable evidence‑based strategic decisions on the allocation of financial, human, and infrastructure resources, thereby improving the quality of studies. This thesis analyses the overall admission rates of Vilnius University and its individual faculties for the period 2000-2027, applying AR, ARIMA, and Holt’s exponential smoothing models. In addition, correlation analysis is used to assess the impact of demographic factors and to determine whether this effect on admission outcomes is statistically significant. The study found that AR, ARIMA, and Holt’s exponential smoothing methods are suitable for forecasting Vilnius University admission rates. The resulting forecasts show different short-term trends, while the overall university admission rate remains stable. A statistically significant relationship was also identified between the number of admitted students and the number of high school graduates in 9 out of the 12 analysed faculties. |
