The Art and Science of Predicting Enrollment Ed Mills Associate Vice President for Student Affairs Enrollment and Student Support Harres Magee Enrollment Analyst
Enrollment Management is both Art and Science Enrollment managers, committees and analysts often develop annual enrollment prediction models using established trends. But what do you do when environmental factors change so quickly and so radically that your predictions need to change just as fast?
This Session To help, we will offer insights on: Reviewing and adjusting the main external and internal variables that enrollment managers track. Identifying additional data to help you more accurately predict and influence your enrollment. Developing tools that will help you change your enrollment predictions as quickly as the environment changes
Typical Enrollment Funnel
Building the Funnel 18 Month Cycle Step 1 Identify the Goals and Target Step 2 Measure Attrition and Graduation Step 3 Measure Continuation Predict the Continuation Rates Step 4 Calculate the Delta to Target Step 5 Build Recruitment Strategy Predict the Yields
Typical Enrollment Growth Model Admission /Re-Entry Admission /Re-Entry Admission /Re-Entry Attrition/ Graduatio n Attrition/ Graduatio n
One Definition for Enrollment Management "Enrollment management is the collaborative use of targeted strategies throughout the campus which are designed to meet the University s goals for student recruitment, retention and graduation."
Suddenly your analyst says: all our models need to change
"Enrollment management is the collaborative use of targeted strategies throughout the campus which are designed to meet the University s goals for student recruitment, retention and graduation."
"Enrollment management is the collaborative implementation of a strategic enrollment cycle designed to quickly adapt to changing environmental factors to meet the University s goals for student recruitment, retention and graduation."
Strategic Enrollment Management Cycle
Growing Enrollment Admission /Re-Entry Admission /Re-Entry Admission /Re-Entry Attrition/ Graduatio n Attrition/ Graduatio n
Maintaining Enrollment New Student New Student New Student Enrollment (with continuing students) Attrition Attrition Attrition
Decreasing Enrollment New Enrollme nt New Enrollme nt Enrollment (with continuing students) Attrition Attrition Attrition
How do we keep our enrollment models Strategic and adapt them to sudden changes?
Where are Changes Happening? Orders to achieve new targets? Adjustments to Institutional Capacity? Fluctuations in the Recruitment Market? Changes Institutional Enrollment Flow
Target Schizophrenia? Situation In 2009-2010 Academic Year our Chancellor s Office (CO) requested we maintain static enrollment (though demand was increasing) The 2010-2011 AY planning the CO targets were reduced by -10.8%. By September, the CO received funding to increase target by +2.7%. At the end of September the CO increased the target by +1.1%. Enrollment Model In Mid- October the CO again increased enrollment target by +3.9% By the end of that same Mid- October day, the CO again increased target by an additional +2.9%. Finally, for the 2011-12, The CO dropped our coming year target to meet anticipated state 2012 funding by 13 5.9%????
Carpel Tunnel Capacity Syndrome? CSU targets are based on FTES. As such, small changes in AUL can create large variations in FTES. If we want to drive FTES up, we might need to bring headcount down.
Restrict, Restrict, Restrict SPEND! One-time funding in 2010-11 suddenly provided a mechanism to create many new courses for fall 2011 increasing capacity (after having spent a year decreasing capacity). The CSU had already created a graduation initiative and developing many new courses fell in-line with objectives to facilitate graduation.
Drive Those FTES Up! Avg. Unit Load 12 10 8 6 4 2 0 Fall 2011 we hit a new AUL record thanks to one-time funding allowing for added sections
Outcome for Fall 2011 Resident FTES (to target) = 23,109 Resident FTES Target = 21,625 Difference = +1,484 Outcome = +6.86% (with a slightly smaller headcount than initially projected 28,019)
Market Fluctuations Good enrollment managers keep a close watch on their primary recruitment markets (headcount and mix of students). But where are you getting the data and how valid is it?
Market High School Graduation by State (California) 12 10 8 6 4 2 0
Market 12 High School Graduation by Local Region (Sac. Region) 10 8 6 4 2 0
Check the Expiration Date Projections change
Enrollment Flow Famine, Flood or Holding Your Own Enrollment managers often become masters at predicting yields, retention, continuation, progression, etc.
Measuring Continuation Total Enrollment Continuing Students Continuation Rate Delta (fall to fall & Spring to Spring) Fall 2008 29,022 - - - Spring 2009 26,977 23,698 81.66% - Fall 2009 29,249 20,398 75.61% - Spring 2010 25,638 25,230 86.26% 4.60% Fall 2010 27,035 19,355 75.49% -0.12% Spring 2011 26,491 23,880 88.33% 2.07% Fall 2011 28,019 19,967 75.37% -0.12%
Enrollment Revelations? What do you do when your projection models need to be more based on the future than the past?
EM Pop Quiz Can a continuation rate ever be more than 100%
EM Pop Quiz Can a continuation rate ever be more than 100% YES! But how?
Always Disaggregate, Evaluate and Re-compute Spring 2011 CSUS had 8,748 Seniors Fall 2011 CSUS continued 8,836 Seniors That s 101% continuation rate how did we accomplish that?
One Possible Answer: Scenario Building Build tools that allow you to change each of your key variables as the strategic enrollment cycle changes. Hunt for Red October Example
References and Tools
Build Scenarios on All Key Enrollment Variables Admissions Funnels Mix (race/ethnicity, age, gender, 1st Generation, other key factors related to your campus) Retention and Progression Rates Continuation (flow) rates Graduation Funnels Attrition (drop-out, stop-out, transfer out and fail out) Rates Average Unit Loads (disaggregate by level) Others?
Epiphalations The first step to better times is to imagine them Trust your intuition If you think its right, it probably is. If you think its wrong Prepare for the unexpected prediction is really the art of successful gambling Good luck is the result of good planning Getting it right the first time is luck, getting it right thereafter is