This page uses extracted PDF text. Text can misread numbers, dates or table columns. Check the PDF before quoting or interpreting it. Redacted areas remain redacted.
Gaskin and Associates Consulting Group, LLC
NNPS proposal
October 8 2026
33
Appendix E Research and transfer to Newport News
Evidence
What it supports
Local limit
Wake County magnet
lottery study, 2025 [S20]
Attendance, progression and stability
benefits in the studied applicant
population.
Third-grade test effects not robust; no
guarantee that a new theme fills an
NNPS school.
Boston implemented
routing, 2019 [S21]
Reported removal of 50 buses and
about $5m annual saving while
maintaining mean rides.
Separate implemented routing from an
unimplemented bell proposal. Local
guarantees and geography differ.
Dallas ACE staffing bundle,
2023 [S22]
Qualified staff, leadership and
sustained support can improve
outcomes.
Bundle effect, not an isolated stipend
effect; incentives ending affected
retention and results.
Closure receiving
effectiveness, 2012 [S23]
Receiving-school quality shapes
displaced-student outcomes.
Author institutional abstract reviewed,
not the full paper.
Chicago closures, 2018
[S24, S28]
Persistent mathematics losses and
transition burdens are a warning.
Large closure episode differs from
NNPS; receiver resources and
transition support matter.
External results identify approaches worth testing locally; they do not supply an NNPS effect size or
saving. Record staffing stability, actual participation, transport service and outcomes in the pilot.
Evaluate the sequence from intervention to service change to educational result, including alternative
explanations such as composition, wage changes or housing.
For the Wake County study, lottery-based estimates reported about 4.074 fewer cumulative absences
over 2016–19 and a 7.907 percentage-point higher probability of graduating the magnet elementary
school; these are not annual districtwide forecasts. Boston’s $5m is never inserted into this model.
Dallas ACE and Chicago findings are interpreted with their populations and comparison designs, not
as universal promises.Jump to an indexed page (48 of 48)