Enrollment Forecasting in an Open Admissions Environment.

Developing a model for predicting demand for freshmen requirement courses (from freshmen enrollees and from returned enrollees who failed to complete the course in their previous quarters) is the objective of the Freshmen Requirement Study, now partially completed by Federal City College. Work done so far has essentially validated an initial approach to the problem where past enrollment behavior has been taken as a predictive factor. Analysis of the data specially compiled for the study shows that the more numerous and more recent a student's enrollment has been in the past, the higher his probability of return is. The difference between the summer quarter and other quarters has been documented. A model that fits well the observed data on return probabilities has been constructed and, using it, the effects of the various components of past behavior affecting return probabilities have been measured. Future work will validate the model with more recent data, fortify it, if necessary, with other explanatory variables, and put it into operation, setting up systems for data collection analysis. Appendices discuss mathematics of probabilistic predictive models: Operationalizing Predictive Models and Fitting an Additive Model to Logarithm of Reenrollment Probabilities. (Author/JT)