Assessment

How we know learning happened — and make sure the knowing serves the learner. This document is the assessment counterpart to philosophy.md and facilitation.md. It defines the modes of assessment, the mastery scale, and the rules that keep assessment honest, fair, and non-coercive at every grade and in every context.

The measure of this curriculum is not test scores but whether learners become more self-aware, more self-motivated, more capable of joy and belonging. Assessment exists to serve that aim — nothing else.

1. Core stance

Assessment serves the learner, not the other way around. Three commitments flow from that, and every instrument in this document must satisfy all three.

• Mastery-focused, not ranked. A learner is measured against a standard (philosophy.md §8: “Mastery is individual; success is collective”), never against another learner. There is no curve, no percentile, no class rank. Two learners can both be at “mastery” and both be celebrated.
• Growth-oriented, not scored. The question is “how far have you come, and what is next?”, never “how many points did you earn?”. Evidence of growth is recorded, not converted into a verdict.
• The end is flourishing, not scores. The real measure is health and flourishing (philosophy.md §12): does the learner grow more self-aware, self-motivated, and capable of joy, belonging, and care? Assessment is a window onto that, never a substitute for it.

What is forbidden follows directly:

• No high-stakes instruments. No single test that decides a learner’s fate, placement, or worth.
• No score-obsessed instruments. No points-gathering, grade-chasing, or deficit-counting that turns learning into a transaction.
• No ranking instruments. No bell curves, no comparative labels, no public “winners” and “losers.”
• No belief-testing. A learner is never assessed on what they believe, only on what they can do and understand (philosophy.md §11 — inquiry over creed). Big questions are held open; the skill of weighing evidence is assessed, never the conclusion a learner reaches.

2. The five modes and how they interlock

Five modes carry the work of knowing. They are not competing; they interlock, and a healthy unit uses all of them. Each stays non-coercive by being low-stakes (safe to get wrong), private or shared by choice, and used to adjust teaching, not to judge a person.

How they interlock: formative feedback during a lesson shapes practice; a summative check confirms the objective was reached; the performance gives the summative check its real-world shape; the portfolio gathers both into evidence of growth; and self-assessment runs through all of it, so the learner is always a co-author of the record. A unit’s assessment_plan (see §7) names which modes carry which objectives.

3. Teacher / facilitator assessment

The facilitator’s job is to observe what a learner can do, and to say back what they saw — specifically and kindly. Assessment is described in observable terms: verbs like names, points to, explains, builds, asks, connects, corrects, never invisible states like understands or gets it. A rubric is a short list of those observable actions at increasing depth (see Mastery levels below).

• Observe first, judge never. Watch the doing, then describe it back: “I saw you try three ways and find the one that balanced. That’s persistence.”
• No ranking, ever. Feedback names the learner’s own progress (“you did X last week; now you do X and Y”), never their position among peers.
• Specific and kind. Say exactly what worked and one next step. “Your diagram labels the parts correctly — next, show how the parts affect each other.” Never vague praise (“good job”) or vague blame (“careless”).
• Feedback is for growth, not for the record’s sake. In large groups where a teacher cannot observe everyone individually, delegate: peer teaching, self- assessment, and portfolios carry the load (contexts.md).
• Celebrate mastery; never turn it into competition (facilitation.md).

4. Learner self-assessment & mastery tracking

A self-directed learner — alone, or in a room without an adult — must be able to check their own understanding, track mastery of each objective, and know when to move on (facilitation.md, “Facilitating self-directed learners”). The curriculum gives them one simple, universal method.

The mastery checklist

For each objective a learner works on, they keep a checklist with four levels. The levels are the same scale used in rubrics (see Mastery levels), and each is written so the learner can honestly self-place:

Concretely, a learner keeps a per-objective mastery list mapped to the standards tree. It can be a page, a card, or marks in the margin of their portfolio:

Objective D01.S1.K.01 — name basic feelings
  [ ] Not yet        [ ] With help        [ ] On my own        [x] Can teach

Objective D05.S2.02.01 — identify and extend growing patterns and describe how they change
  [ ] Not yet        [ ] With help        [x] On my own        [ ] Can teach

How to self-place honestly. Ask: “Could I do this right now, alone? Could I show it to a friend?” If the answer is no to both, it’s “not yet.” If yes only with a prompt, “with help.” If yes alone, “on my own.” If you could show someone else how, “can teach.” The act of explaining is itself the evidence — “the act of explaining is the learning” (facilitation.md).

When to move on. A learner moves on when a unit’s objectives sit at “on my own” or “can teach.” “With help” and “not yet” are not failures; they are the map of what to revisit — a few more practice rounds, a peer guide, or a return to the procedure. There is no clock and no gate; the checklist is a compass, not a door.

The “teach someone else” check. The strongest self-check is explanation: say it aloud, write it, or teach it to a peer or a younger learner. If the explanation holds up to a real question, the learner knows they know.

5. Bias & fairness checks

Assessment is only fair if it measures the skill, not the learner’s culture, language, or access. Before any assessment is used, run these checks.

• Culture. Does the task presume one culture’s default (a specific holiday, game, food, or “common knowledge”)? Replace it with something every learner can anchor to, or offer choices. Ask “what happens in your home?” rather than “what happens on the holiday we all know?”
• Language. Is the task secretly a reading test? A learner who can do the skill but not read the prompt must still be able to show it — through speaking, drawing, acting, or an oral prompt. Offer response in any language the learner owns.
• Learning style. Is the task reachable through more than one modality — body, voice, image, movement, building? A single written form silently ranks readers over doers. Offer the low-tech path (contexts.md) and let the learner choose the form of their evidence.
• Low-tech / low-resource. Every assessment must run on materials_low_tech (paper, voices, bodies, outdoors). If an assessment requires a device, it is optional enrichment, never the gate.
• Never assess belief. This bears repeating from §1: never grade a learner on what they believe, only on skills and understanding (philosophy.md §11). The skill of weighing accounts is assessable; the account a learner settles on is their own. Big questions — origins, meaning, right and wrong — are held open, always.
• Stress and shame. An assessment that makes a learner feel watched, judged, or afraid measures fear, not learning. Prefer quiet, private, or shared-by-choice formats; let a learner pause or step away (facilitation.md).

6. The observability rule

Nothing is taught without a way to know it was learned.

Every objective must have at least one observable way to be demonstrated. If an objective cannot be observed — if no behavior, product, or explanation could show it — then either the objective is miswritten or it is not an objective at all. This is a hard rule for content authors:

• Every lesson’s objectives: (standard IDs) must map to at least one observable check in its ## Assessment section.
• Every unit’s assessment_plan must cover every objective the unit teaches.
• The rubrics and checklists in §3–§4 are written against those same IDs, so a learner, a facilitator, and a validator can all see which standard each piece of evidence serves.

Assessment is tied to the objectives: IDs, not to a vague “understanding” of a topic. D01.S1.K.01 is assessed by observing a child name or point to a feeling — a concrete, watchable act — never by asking whether they “understand feelings.”

7. Alignment with the schema

Assessment lives in specific fields of the file format (schema.md). Content authors should know exactly where it goes:

• lesson.assessment_type — an array naming which modes the lesson uses (formative, summative, performance, portfolio, self-assessment, or a subset). Declares intent at a glance.
• lesson.objectives — the standard IDs the lesson teaches; the ## Assessment body heading then shows how each is observed (see §6).
• unit.assessment_plan — the unit-level plan naming which modes carry which objectives across the unit, and where the portfolio and self-assessment checkpoints fall.
• scope.assessment_plan — the year-level plan; it should state how the five modes balance across the year and when portfolio reviews happen.
• materials_low_tech / materials_enriched — the material tiers; assessment must be possible on the low-tech tier (§5), with enriched tiers optional.
• context_variants — how assessment changes by context (large-group, multi-age, self-directed, level-grouped, outdoor-only). For example, in a self-directed context the self-assessment checklist is the primary instrument; in multi-age the “can teach someone else” level naturally becomes peer teaching.

The required ## Assessment body heading in every lesson is where the observable check lives, written to the learner (“you”) so a self-directed learner can run it alone (schema.md, “Voice: learner-first”).

8. Age-appropriateness

Assessment looks different at each band, matched to development (development.md). It always stays observable, low-stakes, and free of ranking — but its form matures with the learner.

• K–2 (ages ~5–8). Assessment is observation, play, and portfolio. The facilitator watches what a child does — naming a feeling, taking a turn, building a pattern — and notes it. Portfolios hold drawings, photos, and dictated words. Self-assessment is concrete and immediate (“which face is the happy one?”). No written tests; a “test” at this age is a game the child asks to play again.
• 3–5 (ages ~8–11). Children can do a method (predict, test, observe) on real things, so performance grows: build it, show it, explain it. Checklists stay concrete; learners begin to self-place on the four levels with a prompt. Portfolios gather real products. Feedback stays about the work, not the child.
• 6–8 (ages ~11–14). Metacognition begins, so self-assessment and reflection deepen: “what helped? what would I do differently?” Learners can evaluate sources and weigh evidence — assessment of these skills is about the reasoning shown, not the conclusion reached (inquiry over creed). Portfolios become curated, with the learner choosing their strongest evidence.
• 9–12 (ages ~14–18). Learners are co-authors. Assessment is dominated by self-assessment, reflection, and defended performance: a project, an argument, or a body of work the learner stands behind and explains. The four-level checklist matures into goal-setting and progress tracking against standards (D01.S3.10.01 is literally “direct their own learning … and check their own progress”). The hardest truths are assessed as reasoning under ambiguity, never as a single right answer.

Mastery levels

The single, non-competitive scale used everywhere — in rubrics, in the self- assessment checklist, and in facilitator feedback. Four observable levels, named in the learner’s own voice:

Notes:

• Levels describe mastery of one objective, never a person. A learner can be “on my own” for one objective and “not yet” for the next — that is normal and expected.
• “Can teach someone else” is the deepest level, not an extra burden. Teaching is the clearest evidence of understanding, and it doubles as the cooperative structure the curriculum is built on (contexts.md, peer teaching).
• There is no “failing” level. “Not yet” is an honest, kind, and temporary state.

Worked example: D01.S1.K.01

Objective (from the standards tree): Name and identify basic feelings (happy, sad, angry, scared, calm) in self and others. (D01.S1.K.01, Domain 1 — Self & Identity, Strand 1 — Emotional Literacy.)

“I can” statement (written to the learner): “I can name the feelings happy, sad, angry, scared, and calm, and I can point to them in a picture, a friend, or myself.”

Observable check (the ## Assessment for this objective): give the child a set of face cards, drawn faces, or real faces and ask, “Which one is happy? Which one is scared? Show me a face you made when you felt angry.” The child points, names, or shows — all three count as evidence, so a child who cannot yet read or who speaks a different language can still demonstrate the objective. Low-tech, body-only, no device.

Self-assessment check (the same objective on a learner’s mastery list):

D01.S1.K.01 — name basic feelings
  Not yet    With help    On my own    Can teach
                                      [x]  ← "I named all five, then showed my little
                                            sister which face is scared."

The child reaches “can teach someone else” when they can, without prompting, name the five feelings and help someone else identify them. This is observable, low- stakes, culture- and language-flexible, and maps exactly back to the standard ID — assessment serving the learner, in the smallest possible compass.