Evidence Wing · Exhibit V
Methodology
How the conclusions in this investigation were reached — and the point at which they stop.
Procedure
- 01
Public records first
The investigation began with evidence anyone can retrieve: state assessment reporting, district scorecards, the Education Recovery Scorecard, and federal data on staffing and spending. No school was selected because we already knew it.
- 02
Narrowing by result, not by reputation
Districts entered the inquiry because their published outcomes moved, not because they had been recommended, awarded, or profiled. That ordering matters: it means the practices were found after the results, not selected to explain them.
- 03
Observation before interpretation
In each site, what was recorded first was what could be seen and heard — calendars, posted learning intentions, the language teams used about student work. Field notes describe conditions. They do not assign causes.
- 04
Corroboration across sites
A practice observed in one building is a note. A practice observed independently across unrelated buildings, in different states and different demographics, is a pattern. Only the second is written into the Manifesto.
- 05
Research as naming, not as proof
Existing research — Hattie's syntheses, Timperley, Wiliam, Fullan & Quinn, TNTP — is used to name what was observed and to place it in a wider body of work. It is never used to claim that what happened in one district will happen in yours.
How language is matched to evidence
Multiple witnesses
“Schools consistently find…”
Observed independently in several unrelated sites and supported by published research.
One witness
“In one district…”
Documented carefully in a single setting. Instructive, not generalizable.
Self-report only
“Your reflections suggest…”
What people told us about their own experience, treated as perspective rather than measurement.
No evidence
—
Nothing is written.
Where the conclusions stop
- This is not a controlled study. No school was randomly assigned to anything, and no counterfactual exists for any district described here.
- Improvement in a district is not attributed to a single practice. Where several conditions appear together, they are reported together.
- One district is one witness. Heatly is the first witness, not the general case; corroborating districts are named as corroborating, never merged into a composite.
- Where evidence comes only from what educators told us, the language says so — “Educators described…”, not “Schools do…”.
- Where evidence is absent, nothing is claimed. Silence is a finding we are willing to publish.
No conclusion here claims more than the evidence can faithfully support. Where you believe it does, that is worth telling us.