How teacher mathematics turns into student scores.
Districts have spent a decade buying materials and workshops. The mechanism that actually moves achievement sits somewhere else - in what a teacher understands, and in how they run the conversation in the room.
Districts have spent a decade on math PD. The scores did not follow.
Districts have spent heavily on math professional development for a decade. Over the same period NAEP math results have gone down. That is not a spending problem, and it is not an effort problem.
Most math PD was never designed to change what a teacher does at 9:40 on a Tuesday, so it doesn't. Three patterns show up in almost every district we talk to.
One-day workshops
Research finds professional development below roughly 14 contact hours has no measurable effect on student learning. Most math PD is a single day, and the teacher is back in the room on Monday with the same questions.
Yoon et al., IES 2007New materials, same instruction
Districts adopt, train, and adopt again. The materials are different each cycle. What a teacher asks when a student is stuck, and what happens next to a wrong answer, are not.
Coaching with no model behind it
Feedback without a specific account of what better facilitation looks like turns into preference. It also leaves when the coach does, which is why it rarely survives a staffing change.
The ceiling on a math classroom is the teacher's own mathematics.
Your kids will not do better
than your teachers can do.
A curriculum cannot ask a follow-up question. It cannot tell whether a student has understood something or has recognized the shape of a problem. Only the person in the room can do that, and only as far as their own understanding reaches.
Teacher content knowledge is the single strongest school-level predictor of student math achievement.
Hill, Rowan & Ball (2005) — “Effects of Teachers' Mathematical Knowledge for Teaching on Student Achievement,” American Educational Research Journal. One of the most-cited studies in math education.
Which is why a partnership that starts anywhere other than teachers' own mathematics is starting downstream of the thing that decides the result.
Most students consume mathematics. Yours can produce it.
Two people can reach the same answer by completely different routes. One retrieves a procedure and executes it. The other sees how the numbers are put together. Switch between them below on the same expression.
School mathematics today
The Consumer
- Rewards memorizing disparate procedures
- Rewards algorithm deployment over understanding
- Treats math as a static pile of skills and facts
- Measures only the answer
- Assumes math must be made interesting for the learner
Same problem, this way of seeing
- 8 × 7 = 56, plus 5, is 61.
- √9 = 3. 7 × 8 = 56. Three plus 56 is 59.
- 61 is greater than 59.
Two multiplications, a square root, and about forty seconds.
A teacher cannot grow that second habit in a room until they have done the mathematics that way themselves. That is what the partnership spends its first year on.
A prompt with no procedure to reach for, and no back of the book.
A Thought Exercise is a question the whole room has to reason through together. It is the core move of the Generative Math Learning Framework, and it is what teachers spend the summer immersive doing as learners before they ever facilitate one.
8(7) + 5 ______ √9 + 7 · 8
Compare these expressions. Insert >, <, or =. Convince the room.
There is no shortcut to hide behind here. Teachers have to reason from what the numbers are, defend that reasoning to colleagues who will push back, and then learn to run the same conversation with thirty students.
Generative Math Learning Framework · ©2025 Moore and Lewis
Three things that make it work
Students argue, and change their minds
A claim gets defended, challenged and revised in front of everyone. Students change position because a classmate made a better case, not because the teacher signaled the answer.
Every claim traces back to the numbers
Nothing rests on a remembered procedure. A student who can reason from what the numbers are can rebuild the method whenever they need it; a student who only memorized it stops the moment a problem stops matching the template.
A right answer still has to be defended
The room is the judge, so an undefended correct answer earns nothing and a well-argued wrong one teaches everybody something. Reasoning gets practiced out loud, under pressure, in front of peers.
What this partnership is not.
Three things leaders reasonably assume we are, because most of what gets sold into a math department is one of them.
Not a curriculum adoption
Nothing here replaces your materials. It changes how teachers understand the mathematics inside whatever you already use, and how they run the conversation about it.
Not a coaching subscription
Coaching is part of Year Two, aimed at the teachers you are growing into leaders. It is not the product, and it is not something you keep paying for to hold the gains.
Not a student product
Nothing we build touches your students directly. The rehearsal app is a practice room for teachers, with AI students - your actual students only ever meet their own teacher.
The fastest way to judge this is to do the mathematics yourself.
A 45-minute conversation first · the half-day is priced at cost