Comparison of Student and Class Projection Methods in Indonesia Primary and Secondary Education

Accurate planning in education requires reliable student and class size projections to anticipate resource needs, address challenges and ensure education quality. This study aims to respond to the need to evaluate various projection methods to identify the most effective approach in estimating student and class size in educational institutions in Indonesia, while assessing the effectiveness of eight different methods in projecting student and class size. The method used is comparative research with quantitative analysis techniques. The subjects of this study were the number of students and study groups from the 2019/2020 to 2022/2023 academic years in all provinces in Indonesia, then assessed the accuracy of each projection method for the 2023/2024 academic year and the first semester of 2024/2025, Secondary data collection was obtained through the Ministry of Education and Culture's Dapodik website. Data analysis was conducted through inferential statistical analysis. Accuracy assessment was conducted by calculating the error rate using MAPE to evaluate the effectiveness of each method in forecasting the number of students and study groups. Furthermore, each method was given a score to obtain an order based on the level of accuracy. The results of this study reveal that the order of the most accurate methods of projecting the number of students is Naive Forecasting, Autoregressive Integrated Moving Average, Linear Regression, Simple Exponential Smoothing, Moving Average, Holt's Method, Linear Regression, Autoregressive Integrated Moving Average, Naive Forecasting, Autoregressive, Autoregressive Moving Average, Moving Average, and finally Simple Exponential Smoothing. The results of this research are expected to help anticipate student growth and formulate more efficient, sustainable, and equitable education resource allocation strategies.