Overview of the Detailed Enrollment Prediction Model:

The purpose of the UCF Enrollment Prediction Model is to provide a means of estimating headcount (HC) and student credit hours (SCH) by student classification and semester for a prediction year and five subsequent years. The model is “tuned” using a Base Year in order to predict enrollment for the following year, termed the Prediction Year. The overall flow of the model is illustrated in the below graphics. More detailed illustrations of the undergraduate and the graduate portions are also included.

The model builds the student headcount by starting with the returning Fall students. The undergraduates are estimated using cohort retention from the previous 10 years. Returning graduates are based on the past two-year returning rate. Estimates of new students are added to comprise the estimated Fall enrollment. Spring and Summer enrollments use the previous semester enrollment multiplied by the previous year’s semester transition (continuation) fraction plus the estimated new students for that term. Because the retention and transition parameters can vary, the model uses a set of multiplicative adjustment parameters that are computed so that the model, based on the previous year’s data, “fits” the actual enrollment from the previous year perfectly. The resulting model with the adjustment parameters is then used with current year enrollment and the expected new students to predict the following year enrollment by classification. The predicted headcounts are used to estimate the fundable student credit hours by semester, and the annual SCH are used to estimate the fundable FTE by level.

Because of the observed and anticipated increasing enrollment, UCF has continued to revise and update its 5-year enrollment prediction model. In validation tests using historical data, the model was found to predict headcount accurately within 0.5% for a one-year projection and within about 2% for a five-year projection and predict FTEs within 1% for a one-year projection and within about 4% for a five-year projection. The model was accepted as providing reliable estimates. The detailed enrollment prediction model is currently used for short-term (5-year) enrollment predictions as well as the starting point for longer-term enrollment projections.

UCF Undergraduate Enrollment Prediction Model Details:

Model Components:

New Student Input
The primary input required by the model is the estimated number of new students by type: First Time in College Students (FTICs), Community College Transfers (CCT), Other Transfers (OT), and Graduate Students for each semester over the planning horizon (prediction year plus five subsequent years). The estimated numbers of new FTICs, CCTs and OTs are provided by the Vice President of Marketing, Communications, and Admissions (MCA) and the Executive Director of Admissions, while the estimated numbers of new Graduate Students are provided by the Vice Provost and Dean of Graduate Studies.

Undergraduate Fall Retention Fractions
The model uses cohort-based retention fractions that indicate the observed surviving fraction of undergraduate students from a given annual entering cohort that are enrolled in a given classification in the Fall a specified number of years since initial entry. Ten years of entering cohorts are used to calculate the survival fraction retained in a given classification after one year, after two years, …, after ten years. The model uses a two-year average of the fractions retained after a given number of years. For the Base Year analysis, the one-year retention average is generated by the cohorts three-years prior and two-years prior to the Base Year and continues back one year to calculate all ten years of retention fractions. The Prediction Year analysis uses the average of the two years prior to the Base Year for the one-year retention average.

Graduate Fall Continuation Fractions
For Graduate Students, the model estimates the graduate students continued in the Fall as the number of students in the previous Fall multiplied by the fraction of students from the prior year who continued (two-year average.) This fraction is computed only using the total number of graduate students and not on a cohort analysis.

New Undergraduate Student Allocation Fractions
New undergraduate students for a given type (FTIC, CCT, OT) are allocated to a student classification (Freshman, Sophomore, Junior, Senior) in proportion to the actual allocation in the previous year.

Semester Transition Fractions
Students in a given classification in a given semester are allocated to student classifications in the subsequent semester (Spring to Summer, Fall to Spring) in proportion to their actual “transition” in the corresponding semesters of the previous year. These are added to the new students to obtain the estimated enrollment by classification.