---
name: planning-curriculum
description: Plan what a class gets taught and in what order. Use when a teacher wants to build or rework a unit, lay out a term or year as a scope and sequence, turn a set of standards into something teachable, or tighten a unit whose activities and assessments have drifted apart. Settles the understanding a unit is for, and the evidence that would show it, before any lesson is chosen, schedules review rather than teaching each topic once, and flags any outcome that is taught but not assessed or assessed but not taught. For grades 1 through 8.
---

# Planning curriculum

Most units go wrong in the first ten minutes of planning. The teacher starts from
what she already has — the video, the poster lesson, the group task that went
well last year — puts them in a sensible order, and writes the test on the Friday
before it happens, out of whatever the class turned out to have done. Nothing
about that is lazy. It is the wrong end first, and it produces three busy weeks
after which nobody can say what the children understand that they did not before.

This skill runs the other way. Decide what they should walk out understanding,
decide what would show you they do, and only then pick what happens in the
lessons. It works at two scales — one unit, or a term or a year — and both end
at the same table: every outcome, where it is taught, where it is assessed.

The router (`using-gpa-skills`) has already screened for what does not belong in
a planning conversation, drawn the line at diagnosis, and set how much to carry
about any particular child. Rely on that.

## Why the method, and where to say so

A unit plan gets read by someone other than its author — a department head, a
new colleague picking up the class, a parent asking why this is what her child
did in March. Name the method where it is carrying the weight of a decision, and
nowhere else. Four references do nearly all of that work:

- **Backwards design** — Wiggins and McTighe, *Understanding by Design* (1998;
  second edition 2005). Outcomes first, then the evidence, then the lessons, and
  the assessment settled before a single activity is chosen. This is the shape of
  Mode A below.
- **Constructive alignment** — Biggs and Tang, *Teaching for Quality Learning at
  University* (2011). What is assessed, what is taught, and what was intended
  have to be the same three things. Written about universities, and the argument
  carries down unmodified; the alignment table is where it lands.
- **Learning intentions and success criteria** — Hattie, *Visible Learning*
  (2009). Being clear with children about what they are aiming at and what good
  would look like is among the larger effects in his synthesis, and it costs
  nothing. It is also why a vague outcome gets sharpened here before anything is
  designed.
- **Distributed practice** — Cepeda et al. (2006); Dunlosky et al. (2013). A
  topic met again after a gap is held far longer than a topic met once. This is
  the reason Mode B schedules the coming back, rather than leaving it to whether
  there is time in May. What they establish is the gap. Which pattern of gaps
  wins is not settled, so cite them for the spacing and not for the shape of it.

Cite them in the plan itself in a line or two, not a paragraph.

## Read what the school has already settled

Before asking the teacher anything, read these four:

- `school/gpa-context.md#standards-framework` — what the plan maps to.
- `school/gpa-context.md#grade-bands` — what the school calls the grouping, so
  the plan uses the school's words rather than yours.
- `school/gpa-context.md#term-calendar` — how long a term actually is.
- `school/gpa-context.md#specials-and-schedule` — how many periods the subject
  gets a week, and what else is competing for the timetable.

**Where any of those is blank, ask.** Do not name a standards framework, do not
assume a term length, do not guess a number of periods a week, and do not put a
date on which reports go home into a plan that was not given one. A scope and
sequence built on an invented calendar is worse than none, because it looks
finished: the teacher discovers in week nine that the plan assumed a week she
never had, and by then four units have been rewritten around it.

**Where the anchor is blank and the teacher cannot answer either, most of it
degrades rather than stops.** If nobody can say which framework the school
teaches to, design in general terms, say in the plan that you did, and leave the
mapping empty rather than filling it with a framework nobody chose. Two things
do not degrade: the grade, and how long the unit runs. Ask again for those and
say why, rather than designing around a guess at them.

## Gather the rest in one message

Four things, asked together, not as an interview:

1. **The topic, or the outcomes.** Standards, a unit title, or a sentence about
   what she wants them to get.
2. **The grade.** This one changes everything below and is not optional.
3. **How long.** Weeks, or lessons, and how many lessons a week if the schedule
   anchor did not say.
4. **One unit, or a sequence.** This picks the mode.

Anything else — a text she has to use, a district assessment she has to give,
what went wrong last year, what this class already cannot do — is worth having
and is not worth blocking on. Ask for it in the same message and design without
it if it does not come.

## Sharpen the outcome before designing anything

"Cover the Great Lakes" is not an outcome. It is a topic with a verb in front of
it, and a unit built on it can only be assessed by asking what the children
happen to remember.

Turn it into three lists before Stage 1:

- **Know** — the facts, the names, the vocabulary. Recallable.
- **Understand** — the principle underneath, stated as a sentence a child could
  disagree with. Transfers past this unit.
- **Do** — what they can perform. Observable, or it is not a Do.

For the Great Lakes, in grade 5: know the five lakes by name and position, and
know the words watershed, port, channel and invasive species. Understand that
where a place sits on the water decides what gets built there, what work people
do, and what they end up arguing over. Do: explain, using one piece of evidence,
why a named lakeshore city grew where it did rather than ten miles up the
shore.

**Then say plainly that you did this.** Show the teacher the sentence she gave
you and the three lists you turned it into, before you build on them. Never
substitute your own outcome silently — a unit designed
against an outcome the teacher did not choose is worse than a vague one, because
she will not notice until it is taught.

## Mode A — one unit

Three stages, in this order. The order is the method; running Stage 3 before
Stage 2 is what produces the misaligned unit this skill exists to prevent.

### Stage 1 — what they should walk out with

**Enduring understandings.** Two or three, written as full sentences that state
a principle rather than a fact. "The Nile flooded every year" is a fact.
"Settlements grow where the things people need arrive reliably" is a principle,
and it is still true in a unit about Detroit. If the sentence could be marked
right or wrong from a single lookup, it belongs in Know.

**Essential questions.** One or two, and they have to stay open. The test is
whether a child could give a better answer in the last week than in the first
without either answer being wrong. "Which of the lakes is deepest?" fails it.
"Why is the city here and not ten miles up the shore?" passes. In grades 1–4 the
question also has to be sayable by the child: "Whose half is bigger?" works;
"How is equivalence represented across models?" is a question for the plan
document, not for the wall.

**Know** and **Do**, from the sharpening step above.

### Stage 2 — the evidence, written before the lessons

Write this before you write a single activity. Not as a discipline exercise —
it is that the assessment is the only definition of the outcome that is precise
enough to plan against.

**One performance task** that requires transfer. The child has to use what was
taught on a situation she has not already been walked through. A task she can
finish by remembering the worked example from Tuesday tells you the example was
memorable, which is not the outcome. State four things: the situation, what the
child produces, which Stage 1 outcomes it is evidence for, and what a good one
looks like in language the child could be shown.

**Other evidence**, and every piece of it named against a specific outcome. Exit
tickets, a hinge question in the middle of a lesson, one thing you watch for
while walking the room, a short quiz. If you cannot say which outcome a piece of
evidence is for, it is not evidence; it is an activity that produces paper.

Naming the evidence is where this stops. Building it — the items, the key, the
rubric — is `writing-assessments`.

### Stage 3 — the lessons

Now sequence, and sequence toward the task in Stage 2. Three habits do most of
the work:

- **Every lesson points at the assessment.** For each one, write the half-line
  that says what it builds toward. A lesson with no such half-line is either
  mis-sequenced or should not be in the unit; say which.
- **Teach the knowledge before the lesson that applies it.** A child cannot
  reason with a term she met eleven minutes ago.
- **Model it, then do it together, then let go.** Independence is the end of the
  sequence, not its starting condition. Name where in the sequence the scaffold
  comes off, because the scaffold nobody removed is the reason the performance
  task got done and nothing was learned.

## Mode B — a term or a year

A scope and sequence, not a list of topics with weeks next to them. Five things
make it a sequence:

1. **Units in order, with durations**, adding up to the term the calendar anchor
   actually gives you. If the anchor was blank and the teacher told you the
   length, say in the plan that the length came from her.
2. **What each unit depends on.** Name the link, not just the order: "regrouping
   needs place value secure, because a child who does not know what the 1 is
   worth is copying a procedure." Dependencies you can state are dependencies a
   teacher can rearrange around; an order with no stated reason gets shuffled the
   first time a snow day lands in it.
3. **Where a recurring skill gets harder.** The same skill appearing in four
   units is not progression unless the demand rises. Say what rises: describe,
   then explain, then compare two explanations and say which is better.
4. **One outcome line per unit, and the week its evidence lands.** Not the
   unit's full Know / Understand / Do — one sentence on what a child can do at
   the end that she could not at the start, and when the thing that shows it
   happens. Without that pair the sequence cannot be checked at all, and is back
   to being a list of topics with weeks next to them.
5. **Scheduled revisiting.** Below.

### Scheduled revisiting is the part that gets cut

Put the return visits in the plan as dated items with a length: the first within
a week or so of finishing the unit, the next a few weeks later, the next around
a month or more after that. Five minutes each. If it is not in the plan with a
week number against it, it will not happen — May is when the plan is furthest
behind and the review was the only thing with no deadline attached.

Those gaps widen for convenience — it spreads four returns across a term without
bunching them — and not because widening is established. The evidence supports
the gap; whether widening gaps beat even ones is argued about, and a teacher who
prefers three evenly spaced returns is on ground just as good. Do not present the
shape as a finding. No gap and no date is the only indefensible version.

Two rules about what goes in those five minutes:

- **It has to be retrieval.** The children answer from memory — three questions
  on whiteboards, a blank diagram to fill, write down everything you can
  remember about X. Re-teaching a topic is not revisiting it, and it costs a
  full lesson instead of five minutes.
- **It has to be low-stakes.** The point is the effort of remembering, not a
  mark. A graded review turns a five-minute retrieval into an event the class
  needs preparing for, and then it gets cut too.

## The alignment table — both modes

Every plan ends with it. Three columns and a verdict: the outcome, where it is
taught, where it is assessed.

**The rows are whatever the plan's grain is.** In Mode A that is each Stage 1
outcome against lessons and pieces of evidence. In Mode B it is each unit's
outcome line against the weeks it is taught and the week its evidence lands —
coarser, and it catches a different failure: not a lesson that assesses nothing,
but a whole unit that does.

The table is not a formality, because it fails in two directions and both are
common:

- **Taught but not assessed.** Two weeks of lessons produce no evidence
  whatsoever. The teacher finds out in June that she does not know whether it
  landed. Fix by adding one piece of evidence, or — often the honest answer — by
  cutting the outcome, because two weeks was more than it was worth.
- **Assessed but not taught.** The performance task asks for something the
  lessons never built. This one is worse: children are marked down for a gap the
  plan made. Fix by adding the teaching or cutting the demand from the task.

State both directions explicitly, list the specific rows that trip either one,
and say which fix you recommend. A table with a tick in every row and no comment
is a table that was written by filling it in rather than by checking.

## Grades 1–8: what the plan has to fit inside

This is where a plan borrowed from a secondary scheme quietly fails. The design
is fine; it does not fit the children.

**Concrete, then pictorial, then abstract.** Children meet an idea in the hands
first, then as a picture of the thing, then as a symbol. Strips of paper, then a
drawn bar, then 2/4 = 1/2. The sequence traces back to Bruner's modes of
representation (1966), and in grades 1–5 skipping a step does not save a lesson — it
buys a unit where the symbols get manipulated correctly and mean nothing. The
older the class, the shorter the concrete stage, but in grades 6–8 a new idea
still gets one.

**How much fits in a lesson.** Planning defaults, not findings — the teacher in
front of the class knows better, so offer these and use her corrections:

| This heuristic's own bands, not the school's | One lesson can carry | Independent stretch before attention is the variable | New words per unit |
| --- | --- | --- | --- |
| Grades 1–2 | one new idea, met in the hands first | 10–15 minutes | 3–5, each met four or five times |
| Grades 3–5 | one new idea plus a procedure that uses it | 20–25 minutes | 6–10 |
| Grades 6–8 | an idea, a procedure, and the vocabulary that names it | 30–40 minutes | 10–15 |

Two things about it. **The groupings are its own**, invented to hold three sets
of numbers, and are not what the school calls anything; once
`school/gpa-context.md#grade-bands` is filled, use the school's names and let the
numbers fall across them. **And it is your sanity check on a draft, not a thing
to reproduce in the plan** — a school running a workshop model will read 10–15
minutes as wrong, one with a vocabulary programme will read 3–5 words as thin,
and both would be right about their own children.

The middle column is the one that gets ignored. A forty-minute independent task
in grade 2 measures how long a seven-year-old can sit, and you will read the
result as though it measured fractions.

**Vocabulary load.** A term the child cannot decode is a term she cannot think
with. In grades 1–3, check the words in the task are ones she can read before
deciding it assesses the outcome, and count new words across the unit rather than
per lesson — that is the number that gets out of hand.

**Watch the writing tax.** In grades 1–3, and for plenty of children well past
that, a performance task judged on a written explanation measures handwriting,
spelling and sentence construction alongside the outcome. Where the outcome is
not writing, say so and offer the same demand in another form: explain it out
loud to a partner, build it and talk me through it, draw it and label it. That
is not a lower bar. Lowering the bar would be asking for less thinking.

**Say when the plan is too ambitious for the age.** Name the specific thing, and
offer the trade rather than the refusal:

> Three weeks in grade 3 will not carry equivalence, comparison and adding
> fractions. Equivalence alone takes most of it if they are folding and drawing
> before they are writing. I would keep equivalence and comparison, move adding
> to the next unit, and take the freed lessons for the concrete stage.

Producing the over-full plan anyway, with a note that it is tight, is not
honesty. The teacher will teach it, run out of time in week three, and cut the
part at the end — which is the assessment.

## Differentiation, and where it stops being this skill

Differentiation belongs in the plan when it serves the whole class: the same
outcome, reached by different routes. Design it in at Stage 3, not on the Sunday
night before the lesson, and vary only three things — what the children work
with, how they get at it, and how they show what they know. Those three levers
are Tomlinson's (2001; 2014) content, process and product, and so is the warning
about the two ways they get misused. The outcome does not move. When it moves, you have written a second plan and called it a
differentiated one.

Two failures worth naming in the plan itself:

- **Easier work for the children who are behind.** Less demand is not a route to
  the outcome, it is a different destination, and it compounds.
- **More work for the ones who finish first.** Ten more of the same questions is
  a punishment for being quick. Depth is the extension: a harder case, a second
  representation, defend the answer against an objection.

**Where the plan is being built around particular children — three who cannot
regroup, one reading three years ahead — that is `supporting-students`, not
this skill.** Say so and route there. The distinction is what the plan is built
around, not how many names are in it: rebuilding the fractions unit because it
landed for nobody is this skill even though every child is affected. Where a
teacher wants both, say so and take the unit first, because the individual plans
are built on top of what the class is doing.

## A worked unit — Mode A, grade 4, equivalent fractions

The scenario came from the teacher: three weeks, four math lessons a week, and
a class that can partition shapes but reads 1/3 and 1/4 as "three" and "four."
The lesson count and the schedule are hers; nothing about the calendar was
supplied here.

**Understanding.** The same amount can be written in more than one way, because
when the pieces get smaller you need more of them. And: a fraction is only an
amount once you know what it is a fraction of.

**Essential question.** "Whose half is bigger?" — open all three weeks, because
the answer changes the moment the two wholes are different sizes.

**Know.** Numerator and denominator, and which one counts and which one names.
Halves, thirds, quarters, sixths, eighths. That a fraction is a number with a
place on the line, not only a piece of something.

**Do.** Fold or draw a whole into equal parts and name them. Show one amount as
two different fractions, with a picture and with symbols. Place halves, quarters
and eighths on a number line. Decide which of two fractions is more and say how
they know.

**Performance task — the two trays.** Two children each say they got more. One
had half of a small tray of brownies; the other had a third of a large one. The
child gets the two paper trays, decides who is right, and explains it with a
drawing and two sentences. It transfers: the unit taught equivalence and
comparison on a fixed whole, and this changes the whole, which is the thing the
essential question has been circling since day one. Good looks like: the
drawing shows both trays at their real sizes, the explanation says the size of
the whole matters, and the answer is not "a half is always more."

**Other evidence.** End of lesson 3, fold a strip and show two names for the
same length (Do: equivalence). Hinge question in lesson 6 — "is a third always
more than a quarter?" — with whiteboards up, diagnostic rather than evidence:
the class splits, and the split decides how lessons 8 and 9 go. Lesson 9, four fractions
to place on a number line (Do: number line).

**The lessons.** Fold and cut strips, name the parts (concrete). Same strips,
find two names for one length. Draw the strips as bars and write the pair of
symbols underneath (pictorial, then abstract). Build the number line with the
strips laid against it. Compare fractions of the same whole. Compare across
different wholes — the misconception lesson, and the one the task turns on.
Practise the argument in pairs with the success criteria in front of them.
Perform the task. Feed back, and re-open the essential question against the
answers they gave in week one.

**Alignment.**

| Outcome | Taught | Assessed | Verdict |
| --- | --- | --- | --- |
| Understanding 1 — two names, one amount | Lessons 2–4 | Exit ticket L3; performance task | Aligned |
| Understanding 2 — the whole decides | Lessons 8–9 | Performance task | Aligned — see the note on lesson 6 |
| Know — numerator, denominator | Lessons 1–3 | Every written piece requires the words | Aligned |
| Do — compare and justify | Lessons 7–10 | Performance task | Aligned |
| Do — place on a number line | Lessons 5–6 | Lesson 9 check only | Thin |
| Do — explain in writing | Modelled once, lesson 10 | Performance task, both sentences | **Assessed but barely taught** |

Row two first, because it is where a table gets filled in rather than checked.
The hinge runs in lesson 6 and the teaching is in lessons 8–9, so it cannot be
evidence the understanding landed — it comes before there is anything to land.
Its job is to find out who already holds the misconception, which is what decides
how those two lessons go. Left in the Assessed column it ticks the row on one
piece of evidence pretending to be two. The performance task carries that outcome
alone; that is enough, and now it says so.

Then the two flags. The number line is taught across two
lessons and checked once, in a task that never comes back — either the check
carries it, in which case say so in the plan and stop calling it a headline
outcome, or add it to the performance task and drop something else. The written
explanation is the real problem: two sentences are being marked and the class
was shown how to write one, once. Either model it three times with a sentence
frame that comes off by the second, or let the explanation be spoken to the
teacher and mark the drawing. In grade 4 that second option is usually right —
what is being assessed is whether she understands that the whole matters, not
whether she can get it into two sentences.

## A scope-and-sequence fragment — Mode B, the same class, one term

Four units and the returns, over the twelve weeks this teacher said her term ran
to — hers, not a Grosse Pointe Academy term length.

| Weeks | Unit | Depends on | Rising demand |
| --- | --- | --- | --- |
| 1–3 | Place value to 10,000 | — | Read and write numbers |
| 4–6 | Addition and subtraction with regrouping | Place value — a child who cannot say what the 1 is worth is copying steps | Explain a wrong answer |
| 7–9 | Multiplication facts and arrays | Skip counting; arrays reappear as area later | Explain, then choose between two methods |
| 10–12 | Equivalent fractions | Partitioning into equal parts; arrays for the pictures | Compare two explanations and say which is better |

**The returns, scheduled:**

- Week 4, five minutes: three place value questions on whiteboards, cold.
- Week 7, five minutes: place value again, and one regrouping question.
- Week 11, five minutes: regrouping, inside the fractions unit, no warning.
- Week 12, ten minutes: everything, as a blank-page brain dump before her
  reports go home.

The gaps run about a week, then three, then four. That is the scheduling
convenience above rather than a finding, and evenly spaced returns would be
defensible too; what is not negotiable is that each one is retrieval rather than
review. Week 11 is the row that gets deleted first, because by then the fractions
unit is behind. It is the one to defend: regrouping met once in week 6 and never
again is regrouping that will be re-taught from scratch in grade 5.

**Alignment, at term grain.** One line per unit, and the week the evidence lands.

| Unit outcome | Taught | Evidence | Verdict |
| --- | --- | --- | --- |
| Read, write and compare numbers to 10,000 | Weeks 1–3 | End-of-unit check, week 3; retrieval weeks 4 and 7 | Aligned |
| Add and subtract with regrouping, and explain a wrong answer | Weeks 4–6 | End-of-unit check, week 6 — computation only | **Taught but not assessed** |
| Recall the facts, and choose between two methods | Weeks 7–9 | Fluency check week 8; method-choice task week 9 | Aligned |
| Show one amount as two fractions, and compare across wholes | Weeks 10–12 | The two trays, week 12 | Aligned |

The flag is unit 2, and at term scale it is the common one: three weeks teaching
children to explain a wrong answer, and a check that only asks them to get the
right one. Nothing in the term ever asks for the explanation, so in June the
teacher knows who can regroup and not who understands it. Either the week 6 check
gains one item — here is a wrong answer, say what went wrong — or the outcome
loses the explaining clause. The first takes two minutes and is nearly always it.

## How to work

1. Read the four anchors. Ask for what they leave blank, and ask for the four
   things above in the same message.
2. Sharpen the outcome, show the teacher what you sharpened it into, and wait.
3. Pick the mode from what she asked for. Where she wants a unit inside a term
   she has not laid out yet, do Mode B first at low resolution, then Mode A on
   the unit.
4. Run the age check before presenting anything: does a lesson carry what it is
   being asked to carry, is the independent stretch survivable, how many new
   words, and where is the writing tax. Say what you would trade.
5. Build the alignment table last, and read it looking for the two failures
   rather than for ticks.
6. If you can dispatch a subagent, have one read the plan cold and report which
   outcome it thinks the performance task is measuring; otherwise do that pass
   yourself, in sequence, before showing the plan. Where its answer is not the
   outcome you intended, the task is what needs changing.
7. Present it, with what you had to assume and what you would ask next.

If a roster, gradebook or timetable tool is available, use it for the class
list, the marks and the periods; otherwise ask the teacher, and never invent a
name, a mark or a date.

## Handing it back

If you can write files, put the plan there and say where it went — a unit plan is
a document the teacher opens again in October; otherwise present it in the
conversation, whole and ready to paste.

Alongside it, never inside it:

- **What you assumed**, one line each, especially anything you would have read
  from a blank context anchor.
- **The one question that would most improve the plan.** Usually it is what this
  class already cannot do, and it usually rewrites two lessons.
- **Where the plan is thinner than it looks.** The outcome you kept but would
  have cut, the week with no slack in it, the return visit that will be the first
  thing dropped.

A plan that admits where it will break is one a teacher can actually run.