Enrollment Forecasting CHANGES FROM THE OLD METHODOLOGY TO

This is the forecast we see every year in the CIP, and in At-A-Glance

Prior model calculated the impact of development and demographic trends on observable class sizes.

We moved from a Cohort Survival model to one that blends four submodels. OLD

ES Example 4 models are represented by different colored lines. But final projection selectively

HS Example 4 models are represented by different colored lines. But final projection selectively

Additional Comparisons 1200 Enrollemnt Forecast (Students) 900 700 600 500 400 300 200 1200

Primary Concerns Weightings and Validation • Model validation would have consisted of 1) generating

What You Can Do • • Check your forecasts Write to the Bo. E

Slides: 13

Download presentation

Enrollment Forecasting CHANGES FROM THE OLD METHODOLOGY TO THE NEW

This is the forecast we see every year in the CIP, and in At-A-Glance sheets.

Prior model calculated the impact of development and demographic trends on observable class sizes. Cohort Survival Incoming Kindergarteners Approved Development Generation Rates

Prior model – Development

Prior model - Generation Rates

Prior model - Projection

We moved from a Cohort Survival model to one that blends four submodels. OLD METHODOLOGY NEW METHODOLOGY Incoming Kindergarteners Average Percentage Increase Cohort Survival Approved Development Linear Regression Generation Rates Students Per Household

New - Four Weighted Models

ES Example 4 models are represented by different colored lines. But final projection selectively discounts some data.

HS Example 4 models are represented by different colored lines. But final projection selectively discounts some data.

Additional Comparisons 1200 Enrollemnt Forecast (Students) 900 700 600 500 400 300 200 1200 2018 2020 2022 Rachel Carson ES 800 600 400 200 0 2016 2024 2500 1000 2018 2020 2022 2024 Northwood HS 2000 800 1500 600 1000 400 200 0 2016 Neelsville MS 1000 800 0 2016 Enrollemnt Forecast (Students) Ashburton ES Enrollemnt Forecast (Students) 1000 2018 2020 2022 2024 FY 19 CIP Forecast 500 FY 20 CIP Forecast 0 2016 2018 2020 2022 2024

Primary Concerns Weightings and Validation • Model validation would have consisted of 1) generating weights to optimize forecasts of historical enrollment, and then 2) applying those same weights to yield the final projections. • Instead, MGT/MCPS applied new weights subjectively (e. g. to make things “look right, ” or to achieve a “gentler trajectory of growth. ” Housing Growth • MGT identifies High, Low/Medium and No growth areas in their 2018 growth management report. • M-NCPPC provided housing growth rates by cluster ranging from 0% to 4. 2% • In spite of this, MGT applies a flat 1% housing growth assumption for every school in the county. Approved Development • MGT was provided 2016 housing numbers. • Specific approvals since then have not been included, and anything in excess of 1% is missing.

What You Can Do • • Check your forecasts Write to the Bo. E Ask to review and participate Ask for detailed forecasts and assumptions • Email me: CIP@MCCPTA. ORG