Estimating Errors in Student Enrollment Forecasting. AIR Forum 1979 Paper.

The use of data on longitudinal student attendance patterns to determine variances, and hence confidence bounds, on student enrollment forecasts, in addition to finding the forecasts themselves, is demonstrated. The formulation of the enrollment model based on longitudinal student attendance patterns is described step by step, presenting the equations used in the model and some sample data obtained through application of the model. Applications of various equations are demonstrated. The model is analyzed assuming that the new admissions each period follow a Poisson distribution. The enrollment permits the development of statements about the approximate behavior of the conditional distribution of the total number of students in attendance at a given time. It is assumed that entering cohorts contain large numbers of students. Equations relative to this function are presented and discussed. Sample data on the forecasts and confidence intervals for each semester enrollment are presented. The model provides a measure of the error that could be present for each forecast. (SC)