The universal ARIMA model did not fit two series concurrently. For example, the series of teacher demand is a case with multiple factors concurrently in its development process. Previous literature assumed that teacher demand can be forecasted independently through the model ARIMA (autoregressive integrated moving average). In this paper, we applied multivariate ARIMA mode, called ARIMAX, for determining the demand of teachers with the series data of students. We selected the series of data as an example from the Ministry of Education, Taiwan. The cross correlation function and transfer function have been applied to build the fittest ARIMAX model. The ARIMAX analyses indicate that predicting the number of teachers with number of students is better than that only using the universal series data of teacher demand. The implication of these findings may provide an optimal research design for determining the demand of teachers under the changing number of students in elementary schools.
ICIC Express Letters Part B: Applications 11(2), p.129-136