How Education Works | Summative Assessment — How Schools Make Defensible Judgments About Learning

summative assessment is a core education mechanism because schools need to know not only what was taught, but what students can now do, under which conditions, and what should happen next. For readers searching for summative assessment, assessment for learning, assessment of learning, curriculum design, evidence of learning or how education works, this guide focuses on how schools make bounded end-point judgments from sampled performances without pretending a score contains the whole learner. The central mechanism is goal → task → evidence → interpretation → decision → later check.

A score, task or curriculum document is never the learning itself. It is a representation from which teachers, students and institutions make inferences. Good educational design therefore asks whether the representation samples the intended capability, whether conditions are fair enough to interpret, whether uncertainty is acknowledged and whether the resulting decision serves the next legitimate educational purpose.

This world-facing guide develops summative assessment across mathematics, science, reading, vocabulary, English, creative writing, homework, study, teaching, examinations and transfer. Adrian, Jo, Ben, Aisha, Ryan, Mira, Clara and Ethan appear only as fictional eduKateSG residents in invented teaching scenarios. They are not testimonials. The practical frameworks are teaching constructions; official curriculum, examination and institutional rules remain authoritative where applicable.

Quick Read: evidence is useful only when the inference is bounded

Adrian gives a task designed to reveal one capability. Jo records the support conditions and inspects the learner’s reasoning, not only the final mark. The next decision follows the evidence that the task can actually support. A short quiz can reveal a retrieval gap; it cannot by itself establish everything a student knows, values or may later become capable of doing.

A robust loop is define the claim → design the evidence → inspect conditions → interpret cautiously → act proportionately → retest where needed. This protects assessment and curriculum from becoming administrative theatre. Evidence should make education more answerable to learning.

For the wider architecture, begin with How Education Works. Use Learning Goals for the intended capability, Diagnostic Teaching for bottlenecks and Learning Progressions for development across time. This article owns how schools make bounded end-point judgments from sampled performances without pretending a score contains the whole learner.

1. Start with the claim you want to make

Before choosing a test, rubric, project or sequence, state the educational claim. Is the learner expected to retrieve, explain, select, perform, create or transfer? The evidence must fit the claim, and the claim must remain narrower than the evidence can justify.

Summative assessment supports an end-point judgment about sampled performance. Its legitimacy depends on alignment, validity, reliability, fairness and disciplined interpretation of what the result does and does not show.

2. Define the claim

For define the claim, summative assessment begins by stating the educational claim and the evidence that could reasonably support it. Ask what capability is being sampled, what conditions shape performance and what decision will follow. Avoid collecting information simply because it is easy to count or administratively familiar.

In the fictional eduKateSG classroom, Adrian compares the intended learning with the actual demand of the task. Jo records important support conditions and inspects patterns rather than treating one response as a complete portrait. When evidence is ambiguous, the next move is another discriminating task or a narrower claim, not greater certainty in the wording.

The action should remain proportionate. If the evidence reveals a prerequisite gap, repair it. If the task samples too little, gather more evidence. If the curriculum introduces a later demand without an earlier dependency, repair the sequence. If performance is secure only under familiar conditions, add delay or transfer before broadening the conclusion.

Finally, inspect what happens after the decision. Did teaching change? Did the learner improve? Did later performance confirm the inference? Did the curriculum make subsequent learning easier to build? Summative assessment becomes trustworthy when its representations remain correctable by the learner’s return.

3. Learning goal

For learning goal, summative assessment begins by stating the educational claim and the evidence that could reasonably support it. Ask what capability is being sampled, what conditions shape performance and what decision will follow. Avoid collecting information simply because it is easy to count or administratively familiar.

In the fictional eduKateSG classroom, Adrian compares the intended learning with the actual demand of the task. Jo records important support conditions and inspects patterns rather than treating one response as a complete portrait. When evidence is ambiguous, the next move is another discriminating task or a narrower claim, not greater certainty in the wording.

The action should remain proportionate. If the evidence reveals a prerequisite gap, repair it. If the task samples too little, gather more evidence. If the curriculum introduces a later demand without an earlier dependency, repair the sequence. If performance is secure only under familiar conditions, add delay or transfer before broadening the conclusion.

Finally, inspect what happens after the decision. Did teaching change? Did the learner improve? Did later performance confirm the inference? Did the curriculum make subsequent learning easier to build? Summative assessment becomes trustworthy when its representations remain correctable by the learner’s return.

4. Success criteria

For success criteria, summative assessment begins by stating the educational claim and the evidence that could reasonably support it. Ask what capability is being sampled, what conditions shape performance and what decision will follow. Avoid collecting information simply because it is easy to count or administratively familiar.

In the fictional eduKateSG classroom, Adrian compares the intended learning with the actual demand of the task. Jo records important support conditions and inspects patterns rather than treating one response as a complete portrait. When evidence is ambiguous, the next move is another discriminating task or a narrower claim, not greater certainty in the wording.

The action should remain proportionate. If the evidence reveals a prerequisite gap, repair it. If the task samples too little, gather more evidence. If the curriculum introduces a later demand without an earlier dependency, repair the sequence. If performance is secure only under familiar conditions, add delay or transfer before broadening the conclusion.

Finally, inspect what happens after the decision. Did teaching change? Did the learner improve? Did later performance confirm the inference? Did the curriculum make subsequent learning easier to build? Summative assessment becomes trustworthy when its representations remain correctable by the learner’s return.

5. Prior knowledge

For prior knowledge, summative assessment begins by stating the educational claim and the evidence that could reasonably support it. Ask what capability is being sampled, what conditions shape performance and what decision will follow. Avoid collecting information simply because it is easy to count or administratively familiar.

In the fictional eduKateSG classroom, Adrian compares the intended learning with the actual demand of the task. Jo records important support conditions and inspects patterns rather than treating one response as a complete portrait. When evidence is ambiguous, the next move is another discriminating task or a narrower claim, not greater certainty in the wording.

The action should remain proportionate. If the evidence reveals a prerequisite gap, repair it. If the task samples too little, gather more evidence. If the curriculum introduces a later demand without an earlier dependency, repair the sequence. If performance is secure only under familiar conditions, add delay or transfer before broadening the conclusion.

Finally, inspect what happens after the decision. Did teaching change? Did the learner improve? Did later performance confirm the inference? Did the curriculum make subsequent learning easier to build? Summative assessment becomes trustworthy when its representations remain correctable by the learner’s return.

6. Prerequisites

For prerequisites, summative assessment begins by stating the educational claim and the evidence that could reasonably support it. Ask what capability is being sampled, what conditions shape performance and what decision will follow. Avoid collecting information simply because it is easy to count or administratively familiar.

In the fictional eduKateSG classroom, Adrian compares the intended learning with the actual demand of the task. Jo records important support conditions and inspects patterns rather than treating one response as a complete portrait. When evidence is ambiguous, the next move is another discriminating task or a narrower claim, not greater certainty in the wording.

The action should remain proportionate. If the evidence reveals a prerequisite gap, repair it. If the task samples too little, gather more evidence. If the curriculum introduces a later demand without an earlier dependency, repair the sequence. If performance is secure only under familiar conditions, add delay or transfer before broadening the conclusion.

Finally, inspect what happens after the decision. Did teaching change? Did the learner improve? Did later performance confirm the inference? Did the curriculum make subsequent learning easier to build? Summative assessment becomes trustworthy when its representations remain correctable by the learner’s return.

7. Curriculum intent

For curriculum intent, summative assessment begins by stating the educational claim and the evidence that could reasonably support it. Ask what capability is being sampled, what conditions shape performance and what decision will follow. Avoid collecting information simply because it is easy to count or administratively familiar.

In the fictional eduKateSG classroom, Adrian compares the intended learning with the actual demand of the task. Jo records important support conditions and inspects patterns rather than treating one response as a complete portrait. When evidence is ambiguous, the next move is another discriminating task or a narrower claim, not greater certainty in the wording.

The action should remain proportionate. If the evidence reveals a prerequisite gap, repair it. If the task samples too little, gather more evidence. If the curriculum introduces a later demand without an earlier dependency, repair the sequence. If performance is secure only under familiar conditions, add delay or transfer before broadening the conclusion.

Finally, inspect what happens after the decision. Did teaching change? Did the learner improve? Did later performance confirm the inference? Did the curriculum make subsequent learning easier to build? Summative assessment becomes trustworthy when its representations remain correctable by the learner’s return.

8. Task design

For task design, summative assessment begins by stating the educational claim and the evidence that could reasonably support it. Ask what capability is being sampled, what conditions shape performance and what decision will follow. Avoid collecting information simply because it is easy to count or administratively familiar.

In the fictional eduKateSG classroom, Adrian compares the intended learning with the actual demand of the task. Jo records important support conditions and inspects patterns rather than treating one response as a complete portrait. When evidence is ambiguous, the next move is another discriminating task or a narrower claim, not greater certainty in the wording.

The action should remain proportionate. If the evidence reveals a prerequisite gap, repair it. If the task samples too little, gather more evidence. If the curriculum introduces a later demand without an earlier dependency, repair the sequence. If performance is secure only under familiar conditions, add delay or transfer before broadening the conclusion.

Finally, inspect what happens after the decision. Did teaching change? Did the learner improve? Did later performance confirm the inference? Did the curriculum make subsequent learning easier to build? Summative assessment becomes trustworthy when its representations remain correctable by the learner’s return.

9. Evidence

For evidence, summative assessment begins by stating the educational claim and the evidence that could reasonably support it. Ask what capability is being sampled, what conditions shape performance and what decision will follow. Avoid collecting information simply because it is easy to count or administratively familiar.

In the fictional eduKateSG classroom, Adrian compares the intended learning with the actual demand of the task. Jo records important support conditions and inspects patterns rather than treating one response as a complete portrait. When evidence is ambiguous, the next move is another discriminating task or a narrower claim, not greater certainty in the wording.

The action should remain proportionate. If the evidence reveals a prerequisite gap, repair it. If the task samples too little, gather more evidence. If the curriculum introduces a later demand without an earlier dependency, repair the sequence. If performance is secure only under familiar conditions, add delay or transfer before broadening the conclusion.

Finally, inspect what happens after the decision. Did teaching change? Did the learner improve? Did later performance confirm the inference? Did the curriculum make subsequent learning easier to build? Summative assessment becomes trustworthy when its representations remain correctable by the learner’s return.

10. Observation

For observation, summative assessment begins by stating the educational claim and the evidence that could reasonably support it. Ask what capability is being sampled, what conditions shape performance and what decision will follow. Avoid collecting information simply because it is easy to count or administratively familiar.

In the fictional eduKateSG classroom, Adrian compares the intended learning with the actual demand of the task. Jo records important support conditions and inspects patterns rather than treating one response as a complete portrait. When evidence is ambiguous, the next move is another discriminating task or a narrower claim, not greater certainty in the wording.

The action should remain proportionate. If the evidence reveals a prerequisite gap, repair it. If the task samples too little, gather more evidence. If the curriculum introduces a later demand without an earlier dependency, repair the sequence. If performance is secure only under familiar conditions, add delay or transfer before broadening the conclusion.

Finally, inspect what happens after the decision. Did teaching change? Did the learner improve? Did later performance confirm the inference? Did the curriculum make subsequent learning easier to build? Summative assessment becomes trustworthy when its representations remain correctable by the learner’s return.

11. Questioning

For questioning, summative assessment begins by stating the educational claim and the evidence that could reasonably support it. Ask what capability is being sampled, what conditions shape performance and what decision will follow. Avoid collecting information simply because it is easy to count or administratively familiar.

In the fictional eduKateSG classroom, Adrian compares the intended learning with the actual demand of the task. Jo records important support conditions and inspects patterns rather than treating one response as a complete portrait. When evidence is ambiguous, the next move is another discriminating task or a narrower claim, not greater certainty in the wording.

The action should remain proportionate. If the evidence reveals a prerequisite gap, repair it. If the task samples too little, gather more evidence. If the curriculum introduces a later demand without an earlier dependency, repair the sequence. If performance is secure only under familiar conditions, add delay or transfer before broadening the conclusion.

Finally, inspect what happens after the decision. Did teaching change? Did the learner improve? Did later performance confirm the inference? Did the curriculum make subsequent learning easier to build? Summative assessment becomes trustworthy when its representations remain correctable by the learner’s return.

12. Retrieval

For retrieval, summative assessment begins by stating the educational claim and the evidence that could reasonably support it. Ask what capability is being sampled, what conditions shape performance and what decision will follow. Avoid collecting information simply because it is easy to count or administratively familiar.

In the fictional eduKateSG classroom, Adrian compares the intended learning with the actual demand of the task. Jo records important support conditions and inspects patterns rather than treating one response as a complete portrait. When evidence is ambiguous, the next move is another discriminating task or a narrower claim, not greater certainty in the wording.

The action should remain proportionate. If the evidence reveals a prerequisite gap, repair it. If the task samples too little, gather more evidence. If the curriculum introduces a later demand without an earlier dependency, repair the sequence. If performance is secure only under familiar conditions, add delay or transfer before broadening the conclusion.

Finally, inspect what happens after the decision. Did teaching change? Did the learner improve? Did later performance confirm the inference? Did the curriculum make subsequent learning easier to build? Summative assessment becomes trustworthy when its representations remain correctable by the learner’s return.

13. Quiz

For quiz, summative assessment begins by stating the educational claim and the evidence that could reasonably support it. Ask what capability is being sampled, what conditions shape performance and what decision will follow. Avoid collecting information simply because it is easy to count or administratively familiar.

In the fictional eduKateSG classroom, Adrian compares the intended learning with the actual demand of the task. Jo records important support conditions and inspects patterns rather than treating one response as a complete portrait. When evidence is ambiguous, the next move is another discriminating task or a narrower claim, not greater certainty in the wording.

The action should remain proportionate. If the evidence reveals a prerequisite gap, repair it. If the task samples too little, gather more evidence. If the curriculum introduces a later demand without an earlier dependency, repair the sequence. If performance is secure only under familiar conditions, add delay or transfer before broadening the conclusion.

Finally, inspect what happens after the decision. Did teaching change? Did the learner improve? Did later performance confirm the inference? Did the curriculum make subsequent learning easier to build? Summative assessment becomes trustworthy when its representations remain correctable by the learner’s return.

14. Exit ticket

For exit ticket, summative assessment begins by stating the educational claim and the evidence that could reasonably support it. Ask what capability is being sampled, what conditions shape performance and what decision will follow. Avoid collecting information simply because it is easy to count or administratively familiar.

In the fictional eduKateSG classroom, Adrian compares the intended learning with the actual demand of the task. Jo records important support conditions and inspects patterns rather than treating one response as a complete portrait. When evidence is ambiguous, the next move is another discriminating task or a narrower claim, not greater certainty in the wording.

The action should remain proportionate. If the evidence reveals a prerequisite gap, repair it. If the task samples too little, gather more evidence. If the curriculum introduces a later demand without an earlier dependency, repair the sequence. If performance is secure only under familiar conditions, add delay or transfer before broadening the conclusion.

Finally, inspect what happens after the decision. Did teaching change? Did the learner improve? Did later performance confirm the inference? Did the curriculum make subsequent learning easier to build? Summative assessment becomes trustworthy when its representations remain correctable by the learner’s return.

15. Assignment

For assignment, summative assessment begins by stating the educational claim and the evidence that could reasonably support it. Ask what capability is being sampled, what conditions shape performance and what decision will follow. Avoid collecting information simply because it is easy to count or administratively familiar.

In the fictional eduKateSG classroom, Adrian compares the intended learning with the actual demand of the task. Jo records important support conditions and inspects patterns rather than treating one response as a complete portrait. When evidence is ambiguous, the next move is another discriminating task or a narrower claim, not greater certainty in the wording.

The action should remain proportionate. If the evidence reveals a prerequisite gap, repair it. If the task samples too little, gather more evidence. If the curriculum introduces a later demand without an earlier dependency, repair the sequence. If performance is secure only under familiar conditions, add delay or transfer before broadening the conclusion.

Finally, inspect what happens after the decision. Did teaching change? Did the learner improve? Did later performance confirm the inference? Did the curriculum make subsequent learning easier to build? Summative assessment becomes trustworthy when its representations remain correctable by the learner’s return.

16. Project

For project, summative assessment begins by stating the educational claim and the evidence that could reasonably support it. Ask what capability is being sampled, what conditions shape performance and what decision will follow. Avoid collecting information simply because it is easy to count or administratively familiar.

In the fictional eduKateSG classroom, Adrian compares the intended learning with the actual demand of the task. Jo records important support conditions and inspects patterns rather than treating one response as a complete portrait. When evidence is ambiguous, the next move is another discriminating task or a narrower claim, not greater certainty in the wording.

The action should remain proportionate. If the evidence reveals a prerequisite gap, repair it. If the task samples too little, gather more evidence. If the curriculum introduces a later demand without an earlier dependency, repair the sequence. If performance is secure only under familiar conditions, add delay or transfer before broadening the conclusion.

Finally, inspect what happens after the decision. Did teaching change? Did the learner improve? Did later performance confirm the inference? Did the curriculum make subsequent learning easier to build? Summative assessment becomes trustworthy when its representations remain correctable by the learner’s return.

17. Performance task

For performance task, summative assessment begins by stating the educational claim and the evidence that could reasonably support it. Ask what capability is being sampled, what conditions shape performance and what decision will follow. Avoid collecting information simply because it is easy to count or administratively familiar.

In the fictional eduKateSG classroom, Adrian compares the intended learning with the actual demand of the task. Jo records important support conditions and inspects patterns rather than treating one response as a complete portrait. When evidence is ambiguous, the next move is another discriminating task or a narrower claim, not greater certainty in the wording.

The action should remain proportionate. If the evidence reveals a prerequisite gap, repair it. If the task samples too little, gather more evidence. If the curriculum introduces a later demand without an earlier dependency, repair the sequence. If performance is secure only under familiar conditions, add delay or transfer before broadening the conclusion.

Finally, inspect what happens after the decision. Did teaching change? Did the learner improve? Did later performance confirm the inference? Did the curriculum make subsequent learning easier to build? Summative assessment becomes trustworthy when its representations remain correctable by the learner’s return.

18. Examination

For examination, summative assessment begins by stating the educational claim and the evidence that could reasonably support it. Ask what capability is being sampled, what conditions shape performance and what decision will follow. Avoid collecting information simply because it is easy to count or administratively familiar.

In the fictional eduKateSG classroom, Adrian compares the intended learning with the actual demand of the task. Jo records important support conditions and inspects patterns rather than treating one response as a complete portrait. When evidence is ambiguous, the next move is another discriminating task or a narrower claim, not greater certainty in the wording.

The action should remain proportionate. If the evidence reveals a prerequisite gap, repair it. If the task samples too little, gather more evidence. If the curriculum introduces a later demand without an earlier dependency, repair the sequence. If performance is secure only under familiar conditions, add delay or transfer before broadening the conclusion.

Finally, inspect what happens after the decision. Did teaching change? Did the learner improve? Did later performance confirm the inference? Did the curriculum make subsequent learning easier to build? Summative assessment becomes trustworthy when its representations remain correctable by the learner’s return.

19. Rubric

For rubric, summative assessment begins by stating the educational claim and the evidence that could reasonably support it. Ask what capability is being sampled, what conditions shape performance and what decision will follow. Avoid collecting information simply because it is easy to count or administratively familiar.

In the fictional eduKateSG classroom, Adrian compares the intended learning with the actual demand of the task. Jo records important support conditions and inspects patterns rather than treating one response as a complete portrait. When evidence is ambiguous, the next move is another discriminating task or a narrower claim, not greater certainty in the wording.

The action should remain proportionate. If the evidence reveals a prerequisite gap, repair it. If the task samples too little, gather more evidence. If the curriculum introduces a later demand without an earlier dependency, repair the sequence. If performance is secure only under familiar conditions, add delay or transfer before broadening the conclusion.

Finally, inspect what happens after the decision. Did teaching change? Did the learner improve? Did later performance confirm the inference? Did the curriculum make subsequent learning easier to build? Summative assessment becomes trustworthy when its representations remain correctable by the learner’s return.

20. Mark scheme

For mark scheme, summative assessment begins by stating the educational claim and the evidence that could reasonably support it. Ask what capability is being sampled, what conditions shape performance and what decision will follow. Avoid collecting information simply because it is easy to count or administratively familiar.

In the fictional eduKateSG classroom, Adrian compares the intended learning with the actual demand of the task. Jo records important support conditions and inspects patterns rather than treating one response as a complete portrait. When evidence is ambiguous, the next move is another discriminating task or a narrower claim, not greater certainty in the wording.

The action should remain proportionate. If the evidence reveals a prerequisite gap, repair it. If the task samples too little, gather more evidence. If the curriculum introduces a later demand without an earlier dependency, repair the sequence. If performance is secure only under familiar conditions, add delay or transfer before broadening the conclusion.

Finally, inspect what happens after the decision. Did teaching change? Did the learner improve? Did later performance confirm the inference? Did the curriculum make subsequent learning easier to build? Summative assessment becomes trustworthy when its representations remain correctable by the learner’s return.

21. Worked response

For worked response, summative assessment begins by stating the educational claim and the evidence that could reasonably support it. Ask what capability is being sampled, what conditions shape performance and what decision will follow. Avoid collecting information simply because it is easy to count or administratively familiar.

In the fictional eduKateSG classroom, Adrian compares the intended learning with the actual demand of the task. Jo records important support conditions and inspects patterns rather than treating one response as a complete portrait. When evidence is ambiguous, the next move is another discriminating task or a narrower claim, not greater certainty in the wording.

The action should remain proportionate. If the evidence reveals a prerequisite gap, repair it. If the task samples too little, gather more evidence. If the curriculum introduces a later demand without an earlier dependency, repair the sequence. If performance is secure only under familiar conditions, add delay or transfer before broadening the conclusion.

Finally, inspect what happens after the decision. Did teaching change? Did the learner improve? Did later performance confirm the inference? Did the curriculum make subsequent learning easier to build? Summative assessment becomes trustworthy when its representations remain correctable by the learner’s return.

22. Open response

For open response, summative assessment begins by stating the educational claim and the evidence that could reasonably support it. Ask what capability is being sampled, what conditions shape performance and what decision will follow. Avoid collecting information simply because it is easy to count or administratively familiar.

In the fictional eduKateSG classroom, Adrian compares the intended learning with the actual demand of the task. Jo records important support conditions and inspects patterns rather than treating one response as a complete portrait. When evidence is ambiguous, the next move is another discriminating task or a narrower claim, not greater certainty in the wording.

The action should remain proportionate. If the evidence reveals a prerequisite gap, repair it. If the task samples too little, gather more evidence. If the curriculum introduces a later demand without an earlier dependency, repair the sequence. If performance is secure only under familiar conditions, add delay or transfer before broadening the conclusion.

Finally, inspect what happens after the decision. Did teaching change? Did the learner improve? Did later performance confirm the inference? Did the curriculum make subsequent learning easier to build? Summative assessment becomes trustworthy when its representations remain correctable by the learner’s return.

23. Multiple choice

For multiple choice, summative assessment begins by stating the educational claim and the evidence that could reasonably support it. Ask what capability is being sampled, what conditions shape performance and what decision will follow. Avoid collecting information simply because it is easy to count or administratively familiar.

In the fictional eduKateSG classroom, Adrian compares the intended learning with the actual demand of the task. Jo records important support conditions and inspects patterns rather than treating one response as a complete portrait. When evidence is ambiguous, the next move is another discriminating task or a narrower claim, not greater certainty in the wording.

The action should remain proportionate. If the evidence reveals a prerequisite gap, repair it. If the task samples too little, gather more evidence. If the curriculum introduces a later demand without an earlier dependency, repair the sequence. If performance is secure only under familiar conditions, add delay or transfer before broadening the conclusion.

Finally, inspect what happens after the decision. Did teaching change? Did the learner improve? Did later performance confirm the inference? Did the curriculum make subsequent learning easier to build? Summative assessment becomes trustworthy when its representations remain correctable by the learner’s return.

24. Distractors

For distractors, summative assessment begins by stating the educational claim and the evidence that could reasonably support it. Ask what capability is being sampled, what conditions shape performance and what decision will follow. Avoid collecting information simply because it is easy to count or administratively familiar.

In the fictional eduKateSG classroom, Adrian compares the intended learning with the actual demand of the task. Jo records important support conditions and inspects patterns rather than treating one response as a complete portrait. When evidence is ambiguous, the next move is another discriminating task or a narrower claim, not greater certainty in the wording.

The action should remain proportionate. If the evidence reveals a prerequisite gap, repair it. If the task samples too little, gather more evidence. If the curriculum introduces a later demand without an earlier dependency, repair the sequence. If performance is secure only under familiar conditions, add delay or transfer before broadening the conclusion.

Finally, inspect what happens after the decision. Did teaching change? Did the learner improve? Did later performance confirm the inference? Did the curriculum make subsequent learning easier to build? Summative assessment becomes trustworthy when its representations remain correctable by the learner’s return.

25. Sampling

For sampling, summative assessment begins by stating the educational claim and the evidence that could reasonably support it. Ask what capability is being sampled, what conditions shape performance and what decision will follow. Avoid collecting information simply because it is easy to count or administratively familiar.

In the fictional eduKateSG classroom, Adrian compares the intended learning with the actual demand of the task. Jo records important support conditions and inspects patterns rather than treating one response as a complete portrait. When evidence is ambiguous, the next move is another discriminating task or a narrower claim, not greater certainty in the wording.

The action should remain proportionate. If the evidence reveals a prerequisite gap, repair it. If the task samples too little, gather more evidence. If the curriculum introduces a later demand without an earlier dependency, repair the sequence. If performance is secure only under familiar conditions, add delay or transfer before broadening the conclusion.

Finally, inspect what happens after the decision. Did teaching change? Did the learner improve? Did later performance confirm the inference? Did the curriculum make subsequent learning easier to build? Summative assessment becomes trustworthy when its representations remain correctable by the learner’s return.

26. Coverage

For coverage, summative assessment begins by stating the educational claim and the evidence that could reasonably support it. Ask what capability is being sampled, what conditions shape performance and what decision will follow. Avoid collecting information simply because it is easy to count or administratively familiar.

In the fictional eduKateSG classroom, Adrian compares the intended learning with the actual demand of the task. Jo records important support conditions and inspects patterns rather than treating one response as a complete portrait. When evidence is ambiguous, the next move is another discriminating task or a narrower claim, not greater certainty in the wording.

The action should remain proportionate. If the evidence reveals a prerequisite gap, repair it. If the task samples too little, gather more evidence. If the curriculum introduces a later demand without an earlier dependency, repair the sequence. If performance is secure only under familiar conditions, add delay or transfer before broadening the conclusion.

Finally, inspect what happens after the decision. Did teaching change? Did the learner improve? Did later performance confirm the inference? Did the curriculum make subsequent learning easier to build? Summative assessment becomes trustworthy when its representations remain correctable by the learner’s return.

27. Depth

For depth, summative assessment begins by stating the educational claim and the evidence that could reasonably support it. Ask what capability is being sampled, what conditions shape performance and what decision will follow. Avoid collecting information simply because it is easy to count or administratively familiar.

In the fictional eduKateSG classroom, Adrian compares the intended learning with the actual demand of the task. Jo records important support conditions and inspects patterns rather than treating one response as a complete portrait. When evidence is ambiguous, the next move is another discriminating task or a narrower claim, not greater certainty in the wording.

The action should remain proportionate. If the evidence reveals a prerequisite gap, repair it. If the task samples too little, gather more evidence. If the curriculum introduces a later demand without an earlier dependency, repair the sequence. If performance is secure only under familiar conditions, add delay or transfer before broadening the conclusion.

Finally, inspect what happens after the decision. Did teaching change? Did the learner improve? Did later performance confirm the inference? Did the curriculum make subsequent learning easier to build? Summative assessment becomes trustworthy when its representations remain correctable by the learner’s return.

28. Validity

For validity, summative assessment begins by stating the educational claim and the evidence that could reasonably support it. Ask what capability is being sampled, what conditions shape performance and what decision will follow. Avoid collecting information simply because it is easy to count or administratively familiar.

In the fictional eduKateSG classroom, Adrian compares the intended learning with the actual demand of the task. Jo records important support conditions and inspects patterns rather than treating one response as a complete portrait. When evidence is ambiguous, the next move is another discriminating task or a narrower claim, not greater certainty in the wording.

The action should remain proportionate. If the evidence reveals a prerequisite gap, repair it. If the task samples too little, gather more evidence. If the curriculum introduces a later demand without an earlier dependency, repair the sequence. If performance is secure only under familiar conditions, add delay or transfer before broadening the conclusion.

Finally, inspect what happens after the decision. Did teaching change? Did the learner improve? Did later performance confirm the inference? Did the curriculum make subsequent learning easier to build? Summative assessment becomes trustworthy when its representations remain correctable by the learner’s return.

29. Reliability

For reliability, summative assessment begins by stating the educational claim and the evidence that could reasonably support it. Ask what capability is being sampled, what conditions shape performance and what decision will follow. Avoid collecting information simply because it is easy to count or administratively familiar.

In the fictional eduKateSG classroom, Adrian compares the intended learning with the actual demand of the task. Jo records important support conditions and inspects patterns rather than treating one response as a complete portrait. When evidence is ambiguous, the next move is another discriminating task or a narrower claim, not greater certainty in the wording.

The action should remain proportionate. If the evidence reveals a prerequisite gap, repair it. If the task samples too little, gather more evidence. If the curriculum introduces a later demand without an earlier dependency, repair the sequence. If performance is secure only under familiar conditions, add delay or transfer before broadening the conclusion.

Finally, inspect what happens after the decision. Did teaching change? Did the learner improve? Did later performance confirm the inference? Did the curriculum make subsequent learning easier to build? Summative assessment becomes trustworthy when its representations remain correctable by the learner’s return.

30. Fairness

For fairness, summative assessment begins by stating the educational claim and the evidence that could reasonably support it. Ask what capability is being sampled, what conditions shape performance and what decision will follow. Avoid collecting information simply because it is easy to count or administratively familiar.

In the fictional eduKateSG classroom, Adrian compares the intended learning with the actual demand of the task. Jo records important support conditions and inspects patterns rather than treating one response as a complete portrait. When evidence is ambiguous, the next move is another discriminating task or a narrower claim, not greater certainty in the wording.

The action should remain proportionate. If the evidence reveals a prerequisite gap, repair it. If the task samples too little, gather more evidence. If the curriculum introduces a later demand without an earlier dependency, repair the sequence. If performance is secure only under familiar conditions, add delay or transfer before broadening the conclusion.

Finally, inspect what happens after the decision. Did teaching change? Did the learner improve? Did later performance confirm the inference? Did the curriculum make subsequent learning easier to build? Summative assessment becomes trustworthy when its representations remain correctable by the learner’s return.

31. Accessibility

For accessibility, summative assessment begins by stating the educational claim and the evidence that could reasonably support it. Ask what capability is being sampled, what conditions shape performance and what decision will follow. Avoid collecting information simply because it is easy to count or administratively familiar.

In the fictional eduKateSG classroom, Adrian compares the intended learning with the actual demand of the task. Jo records important support conditions and inspects patterns rather than treating one response as a complete portrait. When evidence is ambiguous, the next move is another discriminating task or a narrower claim, not greater certainty in the wording.

The action should remain proportionate. If the evidence reveals a prerequisite gap, repair it. If the task samples too little, gather more evidence. If the curriculum introduces a later demand without an earlier dependency, repair the sequence. If performance is secure only under familiar conditions, add delay or transfer before broadening the conclusion.

Finally, inspect what happens after the decision. Did teaching change? Did the learner improve? Did later performance confirm the inference? Did the curriculum make subsequent learning easier to build? Summative assessment becomes trustworthy when its representations remain correctable by the learner’s return.

32. Accommodations

For accommodations, summative assessment begins by stating the educational claim and the evidence that could reasonably support it. Ask what capability is being sampled, what conditions shape performance and what decision will follow. Avoid collecting information simply because it is easy to count or administratively familiar.

In the fictional eduKateSG classroom, Adrian compares the intended learning with the actual demand of the task. Jo records important support conditions and inspects patterns rather than treating one response as a complete portrait. When evidence is ambiguous, the next move is another discriminating task or a narrower claim, not greater certainty in the wording.

The action should remain proportionate. If the evidence reveals a prerequisite gap, repair it. If the task samples too little, gather more evidence. If the curriculum introduces a later demand without an earlier dependency, repair the sequence. If performance is secure only under familiar conditions, add delay or transfer before broadening the conclusion.

Finally, inspect what happens after the decision. Did teaching change? Did the learner improve? Did later performance confirm the inference? Did the curriculum make subsequent learning easier to build? Summative assessment becomes trustworthy when its representations remain correctable by the learner’s return.

33. Standardisation

For standardisation, summative assessment begins by stating the educational claim and the evidence that could reasonably support it. Ask what capability is being sampled, what conditions shape performance and what decision will follow. Avoid collecting information simply because it is easy to count or administratively familiar.

In the fictional eduKateSG classroom, Adrian compares the intended learning with the actual demand of the task. Jo records important support conditions and inspects patterns rather than treating one response as a complete portrait. When evidence is ambiguous, the next move is another discriminating task or a narrower claim, not greater certainty in the wording.

The action should remain proportionate. If the evidence reveals a prerequisite gap, repair it. If the task samples too little, gather more evidence. If the curriculum introduces a later demand without an earlier dependency, repair the sequence. If performance is secure only under familiar conditions, add delay or transfer before broadening the conclusion.

Finally, inspect what happens after the decision. Did teaching change? Did the learner improve? Did later performance confirm the inference? Did the curriculum make subsequent learning easier to build? Summative assessment becomes trustworthy when its representations remain correctable by the learner’s return.

34. Moderation

For moderation, summative assessment begins by stating the educational claim and the evidence that could reasonably support it. Ask what capability is being sampled, what conditions shape performance and what decision will follow. Avoid collecting information simply because it is easy to count or administratively familiar.

In the fictional eduKateSG classroom, Adrian compares the intended learning with the actual demand of the task. Jo records important support conditions and inspects patterns rather than treating one response as a complete portrait. When evidence is ambiguous, the next move is another discriminating task or a narrower claim, not greater certainty in the wording.

The action should remain proportionate. If the evidence reveals a prerequisite gap, repair it. If the task samples too little, gather more evidence. If the curriculum introduces a later demand without an earlier dependency, repair the sequence. If performance is secure only under familiar conditions, add delay or transfer before broadening the conclusion.

Finally, inspect what happens after the decision. Did teaching change? Did the learner improve? Did later performance confirm the inference? Did the curriculum make subsequent learning easier to build? Summative assessment becomes trustworthy when its representations remain correctable by the learner’s return.

35. Teacher judgment

For teacher judgment, summative assessment begins by stating the educational claim and the evidence that could reasonably support it. Ask what capability is being sampled, what conditions shape performance and what decision will follow. Avoid collecting information simply because it is easy to count or administratively familiar.

In the fictional eduKateSG classroom, Adrian compares the intended learning with the actual demand of the task. Jo records important support conditions and inspects patterns rather than treating one response as a complete portrait. When evidence is ambiguous, the next move is another discriminating task or a narrower claim, not greater certainty in the wording.

The action should remain proportionate. If the evidence reveals a prerequisite gap, repair it. If the task samples too little, gather more evidence. If the curriculum introduces a later demand without an earlier dependency, repair the sequence. If performance is secure only under familiar conditions, add delay or transfer before broadening the conclusion.

Finally, inspect what happens after the decision. Did teaching change? Did the learner improve? Did later performance confirm the inference? Did the curriculum make subsequent learning easier to build? Summative assessment becomes trustworthy when its representations remain correctable by the learner’s return.

36. Scoring

For scoring, summative assessment begins by stating the educational claim and the evidence that could reasonably support it. Ask what capability is being sampled, what conditions shape performance and what decision will follow. Avoid collecting information simply because it is easy to count or administratively familiar.

In the fictional eduKateSG classroom, Adrian compares the intended learning with the actual demand of the task. Jo records important support conditions and inspects patterns rather than treating one response as a complete portrait. When evidence is ambiguous, the next move is another discriminating task or a narrower claim, not greater certainty in the wording.

The action should remain proportionate. If the evidence reveals a prerequisite gap, repair it. If the task samples too little, gather more evidence. If the curriculum introduces a later demand without an earlier dependency, repair the sequence. If performance is secure only under familiar conditions, add delay or transfer before broadening the conclusion.

Finally, inspect what happens after the decision. Did teaching change? Did the learner improve? Did later performance confirm the inference? Did the curriculum make subsequent learning easier to build? Summative assessment becomes trustworthy when its representations remain correctable by the learner’s return.

37. Grades

For grades, summative assessment begins by stating the educational claim and the evidence that could reasonably support it. Ask what capability is being sampled, what conditions shape performance and what decision will follow. Avoid collecting information simply because it is easy to count or administratively familiar.

In the fictional eduKateSG classroom, Adrian compares the intended learning with the actual demand of the task. Jo records important support conditions and inspects patterns rather than treating one response as a complete portrait. When evidence is ambiguous, the next move is another discriminating task or a narrower claim, not greater certainty in the wording.

The action should remain proportionate. If the evidence reveals a prerequisite gap, repair it. If the task samples too little, gather more evidence. If the curriculum introduces a later demand without an earlier dependency, repair the sequence. If performance is secure only under familiar conditions, add delay or transfer before broadening the conclusion.

Finally, inspect what happens after the decision. Did teaching change? Did the learner improve? Did later performance confirm the inference? Did the curriculum make subsequent learning easier to build? Summative assessment becomes trustworthy when its representations remain correctable by the learner’s return.

38. Cut scores

For cut scores, summative assessment begins by stating the educational claim and the evidence that could reasonably support it. Ask what capability is being sampled, what conditions shape performance and what decision will follow. Avoid collecting information simply because it is easy to count or administratively familiar.

In the fictional eduKateSG classroom, Adrian compares the intended learning with the actual demand of the task. Jo records important support conditions and inspects patterns rather than treating one response as a complete portrait. When evidence is ambiguous, the next move is another discriminating task or a narrower claim, not greater certainty in the wording.

The action should remain proportionate. If the evidence reveals a prerequisite gap, repair it. If the task samples too little, gather more evidence. If the curriculum introduces a later demand without an earlier dependency, repair the sequence. If performance is secure only under familiar conditions, add delay or transfer before broadening the conclusion.

Finally, inspect what happens after the decision. Did teaching change? Did the learner improve? Did later performance confirm the inference? Did the curriculum make subsequent learning easier to build? Summative assessment becomes trustworthy when its representations remain correctable by the learner’s return.

39. Rankings

For rankings, summative assessment begins by stating the educational claim and the evidence that could reasonably support it. Ask what capability is being sampled, what conditions shape performance and what decision will follow. Avoid collecting information simply because it is easy to count or administratively familiar.

In the fictional eduKateSG classroom, Adrian compares the intended learning with the actual demand of the task. Jo records important support conditions and inspects patterns rather than treating one response as a complete portrait. When evidence is ambiguous, the next move is another discriminating task or a narrower claim, not greater certainty in the wording.

The action should remain proportionate. If the evidence reveals a prerequisite gap, repair it. If the task samples too little, gather more evidence. If the curriculum introduces a later demand without an earlier dependency, repair the sequence. If performance is secure only under familiar conditions, add delay or transfer before broadening the conclusion.

Finally, inspect what happens after the decision. Did teaching change? Did the learner improve? Did later performance confirm the inference? Did the curriculum make subsequent learning easier to build? Summative assessment becomes trustworthy when its representations remain correctable by the learner’s return.

40. Uncertainty

For uncertainty, summative assessment begins by stating the educational claim and the evidence that could reasonably support it. Ask what capability is being sampled, what conditions shape performance and what decision will follow. Avoid collecting information simply because it is easy to count or administratively familiar.

In the fictional eduKateSG classroom, Adrian compares the intended learning with the actual demand of the task. Jo records important support conditions and inspects patterns rather than treating one response as a complete portrait. When evidence is ambiguous, the next move is another discriminating task or a narrower claim, not greater certainty in the wording.

The action should remain proportionate. If the evidence reveals a prerequisite gap, repair it. If the task samples too little, gather more evidence. If the curriculum introduces a later demand without an earlier dependency, repair the sequence. If performance is secure only under familiar conditions, add delay or transfer before broadening the conclusion.

Finally, inspect what happens after the decision. Did teaching change? Did the learner improve? Did later performance confirm the inference? Did the curriculum make subsequent learning easier to build? Summative assessment becomes trustworthy when its representations remain correctable by the learner’s return.

41. Measurement error

For measurement error, summative assessment begins by stating the educational claim and the evidence that could reasonably support it. Ask what capability is being sampled, what conditions shape performance and what decision will follow. Avoid collecting information simply because it is easy to count or administratively familiar.

In the fictional eduKateSG classroom, Adrian compares the intended learning with the actual demand of the task. Jo records important support conditions and inspects patterns rather than treating one response as a complete portrait. When evidence is ambiguous, the next move is another discriminating task or a narrower claim, not greater certainty in the wording.

The action should remain proportionate. If the evidence reveals a prerequisite gap, repair it. If the task samples too little, gather more evidence. If the curriculum introduces a later demand without an earlier dependency, repair the sequence. If performance is secure only under familiar conditions, add delay or transfer before broadening the conclusion.

Finally, inspect what happens after the decision. Did teaching change? Did the learner improve? Did later performance confirm the inference? Did the curriculum make subsequent learning easier to build? Summative assessment becomes trustworthy when its representations remain correctable by the learner’s return.

42. Conditions

For conditions, summative assessment begins by stating the educational claim and the evidence that could reasonably support it. Ask what capability is being sampled, what conditions shape performance and what decision will follow. Avoid collecting information simply because it is easy to count or administratively familiar.

In the fictional eduKateSG classroom, Adrian compares the intended learning with the actual demand of the task. Jo records important support conditions and inspects patterns rather than treating one response as a complete portrait. When evidence is ambiguous, the next move is another discriminating task or a narrower claim, not greater certainty in the wording.

The action should remain proportionate. If the evidence reveals a prerequisite gap, repair it. If the task samples too little, gather more evidence. If the curriculum introduces a later demand without an earlier dependency, repair the sequence. If performance is secure only under familiar conditions, add delay or transfer before broadening the conclusion.

Finally, inspect what happens after the decision. Did teaching change? Did the learner improve? Did later performance confirm the inference? Did the curriculum make subsequent learning easier to build? Summative assessment becomes trustworthy when its representations remain correctable by the learner’s return.

43. Time limits

For time limits, summative assessment begins by stating the educational claim and the evidence that could reasonably support it. Ask what capability is being sampled, what conditions shape performance and what decision will follow. Avoid collecting information simply because it is easy to count or administratively familiar.

In the fictional eduKateSG classroom, Adrian compares the intended learning with the actual demand of the task. Jo records important support conditions and inspects patterns rather than treating one response as a complete portrait. When evidence is ambiguous, the next move is another discriminating task or a narrower claim, not greater certainty in the wording.

The action should remain proportionate. If the evidence reveals a prerequisite gap, repair it. If the task samples too little, gather more evidence. If the curriculum introduces a later demand without an earlier dependency, repair the sequence. If performance is secure only under familiar conditions, add delay or transfer before broadening the conclusion.

Finally, inspect what happens after the decision. Did teaching change? Did the learner improve? Did later performance confirm the inference? Did the curriculum make subsequent learning easier to build? Summative assessment becomes trustworthy when its representations remain correctable by the learner’s return.

44. Language demand

For language demand, summative assessment begins by stating the educational claim and the evidence that could reasonably support it. Ask what capability is being sampled, what conditions shape performance and what decision will follow. Avoid collecting information simply because it is easy to count or administratively familiar.

In the fictional eduKateSG classroom, Adrian compares the intended learning with the actual demand of the task. Jo records important support conditions and inspects patterns rather than treating one response as a complete portrait. When evidence is ambiguous, the next move is another discriminating task or a narrower claim, not greater certainty in the wording.

The action should remain proportionate. If the evidence reveals a prerequisite gap, repair it. If the task samples too little, gather more evidence. If the curriculum introduces a later demand without an earlier dependency, repair the sequence. If performance is secure only under familiar conditions, add delay or transfer before broadening the conclusion.

Finally, inspect what happens after the decision. Did teaching change? Did the learner improve? Did later performance confirm the inference? Did the curriculum make subsequent learning easier to build? Summative assessment becomes trustworthy when its representations remain correctable by the learner’s return.

45. Technology

For technology, summative assessment begins by stating the educational claim and the evidence that could reasonably support it. Ask what capability is being sampled, what conditions shape performance and what decision will follow. Avoid collecting information simply because it is easy to count or administratively familiar.

In the fictional eduKateSG classroom, Adrian compares the intended learning with the actual demand of the task. Jo records important support conditions and inspects patterns rather than treating one response as a complete portrait. When evidence is ambiguous, the next move is another discriminating task or a narrower claim, not greater certainty in the wording.

The action should remain proportionate. If the evidence reveals a prerequisite gap, repair it. If the task samples too little, gather more evidence. If the curriculum introduces a later demand without an earlier dependency, repair the sequence. If performance is secure only under familiar conditions, add delay or transfer before broadening the conclusion.

Finally, inspect what happens after the decision. Did teaching change? Did the learner improve? Did later performance confirm the inference? Did the curriculum make subsequent learning easier to build? Summative assessment becomes trustworthy when its representations remain correctable by the learner’s return.

46. AI

For ai, summative assessment begins by stating the educational claim and the evidence that could reasonably support it. Ask what capability is being sampled, what conditions shape performance and what decision will follow. Avoid collecting information simply because it is easy to count or administratively familiar.

In the fictional eduKateSG classroom, Adrian compares the intended learning with the actual demand of the task. Jo records important support conditions and inspects patterns rather than treating one response as a complete portrait. When evidence is ambiguous, the next move is another discriminating task or a narrower claim, not greater certainty in the wording.

The action should remain proportionate. If the evidence reveals a prerequisite gap, repair it. If the task samples too little, gather more evidence. If the curriculum introduces a later demand without an earlier dependency, repair the sequence. If performance is secure only under familiar conditions, add delay or transfer before broadening the conclusion.

Finally, inspect what happens after the decision. Did teaching change? Did the learner improve? Did later performance confirm the inference? Did the curriculum make subsequent learning easier to build? Summative assessment becomes trustworthy when its representations remain correctable by the learner’s return.

47. Data

For data, summative assessment begins by stating the educational claim and the evidence that could reasonably support it. Ask what capability is being sampled, what conditions shape performance and what decision will follow. Avoid collecting information simply because it is easy to count or administratively familiar.

In the fictional eduKateSG classroom, Adrian compares the intended learning with the actual demand of the task. Jo records important support conditions and inspects patterns rather than treating one response as a complete portrait. When evidence is ambiguous, the next move is another discriminating task or a narrower claim, not greater certainty in the wording.

The action should remain proportionate. If the evidence reveals a prerequisite gap, repair it. If the task samples too little, gather more evidence. If the curriculum introduces a later demand without an earlier dependency, repair the sequence. If performance is secure only under familiar conditions, add delay or transfer before broadening the conclusion.

Finally, inspect what happens after the decision. Did teaching change? Did the learner improve? Did later performance confirm the inference? Did the curriculum make subsequent learning easier to build? Summative assessment becomes trustworthy when its representations remain correctable by the learner’s return.

48. Privacy

For privacy, summative assessment begins by stating the educational claim and the evidence that could reasonably support it. Ask what capability is being sampled, what conditions shape performance and what decision will follow. Avoid collecting information simply because it is easy to count or administratively familiar.

In the fictional eduKateSG classroom, Adrian compares the intended learning with the actual demand of the task. Jo records important support conditions and inspects patterns rather than treating one response as a complete portrait. When evidence is ambiguous, the next move is another discriminating task or a narrower claim, not greater certainty in the wording.

The action should remain proportionate. If the evidence reveals a prerequisite gap, repair it. If the task samples too little, gather more evidence. If the curriculum introduces a later demand without an earlier dependency, repair the sequence. If performance is secure only under familiar conditions, add delay or transfer before broadening the conclusion.

Finally, inspect what happens after the decision. Did teaching change? Did the learner improve? Did later performance confirm the inference? Did the curriculum make subsequent learning easier to build? Summative assessment becomes trustworthy when its representations remain correctable by the learner’s return.

49. Feedback

For feedback, summative assessment begins by stating the educational claim and the evidence that could reasonably support it. Ask what capability is being sampled, what conditions shape performance and what decision will follow. Avoid collecting information simply because it is easy to count or administratively familiar.

In the fictional eduKateSG classroom, Adrian compares the intended learning with the actual demand of the task. Jo records important support conditions and inspects patterns rather than treating one response as a complete portrait. When evidence is ambiguous, the next move is another discriminating task or a narrower claim, not greater certainty in the wording.

The action should remain proportionate. If the evidence reveals a prerequisite gap, repair it. If the task samples too little, gather more evidence. If the curriculum introduces a later demand without an earlier dependency, repair the sequence. If performance is secure only under familiar conditions, add delay or transfer before broadening the conclusion.

Finally, inspect what happens after the decision. Did teaching change? Did the learner improve? Did later performance confirm the inference? Did the curriculum make subsequent learning easier to build? Summative assessment becomes trustworthy when its representations remain correctable by the learner’s return.

50. Diagnostic teaching

For diagnostic teaching, summative assessment begins by stating the educational claim and the evidence that could reasonably support it. Ask what capability is being sampled, what conditions shape performance and what decision will follow. Avoid collecting information simply because it is easy to count or administratively familiar.

In the fictional eduKateSG classroom, Adrian compares the intended learning with the actual demand of the task. Jo records important support conditions and inspects patterns rather than treating one response as a complete portrait. When evidence is ambiguous, the next move is another discriminating task or a narrower claim, not greater certainty in the wording.

The action should remain proportionate. If the evidence reveals a prerequisite gap, repair it. If the task samples too little, gather more evidence. If the curriculum introduces a later demand without an earlier dependency, repair the sequence. If performance is secure only under familiar conditions, add delay or transfer before broadening the conclusion.

Finally, inspect what happens after the decision. Did teaching change? Did the learner improve? Did later performance confirm the inference? Did the curriculum make subsequent learning easier to build? Summative assessment becomes trustworthy when its representations remain correctable by the learner’s return.

51. Misconceptions

For misconceptions, summative assessment begins by stating the educational claim and the evidence that could reasonably support it. Ask what capability is being sampled, what conditions shape performance and what decision will follow. Avoid collecting information simply because it is easy to count or administratively familiar.

In the fictional eduKateSG classroom, Adrian compares the intended learning with the actual demand of the task. Jo records important support conditions and inspects patterns rather than treating one response as a complete portrait. When evidence is ambiguous, the next move is another discriminating task or a narrower claim, not greater certainty in the wording.

The action should remain proportionate. If the evidence reveals a prerequisite gap, repair it. If the task samples too little, gather more evidence. If the curriculum introduces a later demand without an earlier dependency, repair the sequence. If performance is secure only under familiar conditions, add delay or transfer before broadening the conclusion.

Finally, inspect what happens after the decision. Did teaching change? Did the learner improve? Did later performance confirm the inference? Did the curriculum make subsequent learning easier to build? Summative assessment becomes trustworthy when its representations remain correctable by the learner’s return.

52. Error

For error, summative assessment begins by stating the educational claim and the evidence that could reasonably support it. Ask what capability is being sampled, what conditions shape performance and what decision will follow. Avoid collecting information simply because it is easy to count or administratively familiar.

In the fictional eduKateSG classroom, Adrian compares the intended learning with the actual demand of the task. Jo records important support conditions and inspects patterns rather than treating one response as a complete portrait. When evidence is ambiguous, the next move is another discriminating task or a narrower claim, not greater certainty in the wording.

The action should remain proportionate. If the evidence reveals a prerequisite gap, repair it. If the task samples too little, gather more evidence. If the curriculum introduces a later demand without an earlier dependency, repair the sequence. If performance is secure only under familiar conditions, add delay or transfer before broadening the conclusion.

Finally, inspect what happens after the decision. Did teaching change? Did the learner improve? Did later performance confirm the inference? Did the curriculum make subsequent learning easier to build? Summative assessment becomes trustworthy when its representations remain correctable by the learner’s return.

53. Correction

For correction, summative assessment begins by stating the educational claim and the evidence that could reasonably support it. Ask what capability is being sampled, what conditions shape performance and what decision will follow. Avoid collecting information simply because it is easy to count or administratively familiar.

In the fictional eduKateSG classroom, Adrian compares the intended learning with the actual demand of the task. Jo records important support conditions and inspects patterns rather than treating one response as a complete portrait. When evidence is ambiguous, the next move is another discriminating task or a narrower claim, not greater certainty in the wording.

The action should remain proportionate. If the evidence reveals a prerequisite gap, repair it. If the task samples too little, gather more evidence. If the curriculum introduces a later demand without an earlier dependency, repair the sequence. If performance is secure only under familiar conditions, add delay or transfer before broadening the conclusion.

Finally, inspect what happens after the decision. Did teaching change? Did the learner improve? Did later performance confirm the inference? Did the curriculum make subsequent learning easier to build? Summative assessment becomes trustworthy when its representations remain correctable by the learner’s return.

54. Revision

For revision, summative assessment begins by stating the educational claim and the evidence that could reasonably support it. Ask what capability is being sampled, what conditions shape performance and what decision will follow. Avoid collecting information simply because it is easy to count or administratively familiar.

In the fictional eduKateSG classroom, Adrian compares the intended learning with the actual demand of the task. Jo records important support conditions and inspects patterns rather than treating one response as a complete portrait. When evidence is ambiguous, the next move is another discriminating task or a narrower claim, not greater certainty in the wording.

The action should remain proportionate. If the evidence reveals a prerequisite gap, repair it. If the task samples too little, gather more evidence. If the curriculum introduces a later demand without an earlier dependency, repair the sequence. If performance is secure only under familiar conditions, add delay or transfer before broadening the conclusion.

Finally, inspect what happens after the decision. Did teaching change? Did the learner improve? Did later performance confirm the inference? Did the curriculum make subsequent learning easier to build? Summative assessment becomes trustworthy when its representations remain correctable by the learner’s return.

55. Practice

For practice, summative assessment begins by stating the educational claim and the evidence that could reasonably support it. Ask what capability is being sampled, what conditions shape performance and what decision will follow. Avoid collecting information simply because it is easy to count or administratively familiar.

In the fictional eduKateSG classroom, Adrian compares the intended learning with the actual demand of the task. Jo records important support conditions and inspects patterns rather than treating one response as a complete portrait. When evidence is ambiguous, the next move is another discriminating task or a narrower claim, not greater certainty in the wording.

The action should remain proportionate. If the evidence reveals a prerequisite gap, repair it. If the task samples too little, gather more evidence. If the curriculum introduces a later demand without an earlier dependency, repair the sequence. If performance is secure only under familiar conditions, add delay or transfer before broadening the conclusion.

Finally, inspect what happens after the decision. Did teaching change? Did the learner improve? Did later performance confirm the inference? Did the curriculum make subsequent learning easier to build? Summative assessment becomes trustworthy when its representations remain correctable by the learner’s return.

56. Spacing

For spacing, summative assessment begins by stating the educational claim and the evidence that could reasonably support it. Ask what capability is being sampled, what conditions shape performance and what decision will follow. Avoid collecting information simply because it is easy to count or administratively familiar.

In the fictional eduKateSG classroom, Adrian compares the intended learning with the actual demand of the task. Jo records important support conditions and inspects patterns rather than treating one response as a complete portrait. When evidence is ambiguous, the next move is another discriminating task or a narrower claim, not greater certainty in the wording.

The action should remain proportionate. If the evidence reveals a prerequisite gap, repair it. If the task samples too little, gather more evidence. If the curriculum introduces a later demand without an earlier dependency, repair the sequence. If performance is secure only under familiar conditions, add delay or transfer before broadening the conclusion.

Finally, inspect what happens after the decision. Did teaching change? Did the learner improve? Did later performance confirm the inference? Did the curriculum make subsequent learning easier to build? Summative assessment becomes trustworthy when its representations remain correctable by the learner’s return.

57. Transfer

For transfer, summative assessment begins by stating the educational claim and the evidence that could reasonably support it. Ask what capability is being sampled, what conditions shape performance and what decision will follow. Avoid collecting information simply because it is easy to count or administratively familiar.

In the fictional eduKateSG classroom, Adrian compares the intended learning with the actual demand of the task. Jo records important support conditions and inspects patterns rather than treating one response as a complete portrait. When evidence is ambiguous, the next move is another discriminating task or a narrower claim, not greater certainty in the wording.

The action should remain proportionate. If the evidence reveals a prerequisite gap, repair it. If the task samples too little, gather more evidence. If the curriculum introduces a later demand without an earlier dependency, repair the sequence. If performance is secure only under familiar conditions, add delay or transfer before broadening the conclusion.

Finally, inspect what happens after the decision. Did teaching change? Did the learner improve? Did later performance confirm the inference? Did the curriculum make subsequent learning easier to build? Summative assessment becomes trustworthy when its representations remain correctable by the learner’s return.

58. Retention

For retention, summative assessment begins by stating the educational claim and the evidence that could reasonably support it. Ask what capability is being sampled, what conditions shape performance and what decision will follow. Avoid collecting information simply because it is easy to count or administratively familiar.

In the fictional eduKateSG classroom, Adrian compares the intended learning with the actual demand of the task. Jo records important support conditions and inspects patterns rather than treating one response as a complete portrait. When evidence is ambiguous, the next move is another discriminating task or a narrower claim, not greater certainty in the wording.

The action should remain proportionate. If the evidence reveals a prerequisite gap, repair it. If the task samples too little, gather more evidence. If the curriculum introduces a later demand without an earlier dependency, repair the sequence. If performance is secure only under familiar conditions, add delay or transfer before broadening the conclusion.

Finally, inspect what happens after the decision. Did teaching change? Did the learner improve? Did later performance confirm the inference? Did the curriculum make subsequent learning easier to build? Summative assessment becomes trustworthy when its representations remain correctable by the learner’s return.

59. Memory

For memory, summative assessment begins by stating the educational claim and the evidence that could reasonably support it. Ask what capability is being sampled, what conditions shape performance and what decision will follow. Avoid collecting information simply because it is easy to count or administratively familiar.

In the fictional eduKateSG classroom, Adrian compares the intended learning with the actual demand of the task. Jo records important support conditions and inspects patterns rather than treating one response as a complete portrait. When evidence is ambiguous, the next move is another discriminating task or a narrower claim, not greater certainty in the wording.

The action should remain proportionate. If the evidence reveals a prerequisite gap, repair it. If the task samples too little, gather more evidence. If the curriculum introduces a later demand without an earlier dependency, repair the sequence. If performance is secure only under familiar conditions, add delay or transfer before broadening the conclusion.

Finally, inspect what happens after the decision. Did teaching change? Did the learner improve? Did later performance confirm the inference? Did the curriculum make subsequent learning easier to build? Summative assessment becomes trustworthy when its representations remain correctable by the learner’s return.

60. Learning progressions

For learning progressions, summative assessment begins by stating the educational claim and the evidence that could reasonably support it. Ask what capability is being sampled, what conditions shape performance and what decision will follow. Avoid collecting information simply because it is easy to count or administratively familiar.

In the fictional eduKateSG classroom, Adrian compares the intended learning with the actual demand of the task. Jo records important support conditions and inspects patterns rather than treating one response as a complete portrait. When evidence is ambiguous, the next move is another discriminating task or a narrower claim, not greater certainty in the wording.

The action should remain proportionate. If the evidence reveals a prerequisite gap, repair it. If the task samples too little, gather more evidence. If the curriculum introduces a later demand without an earlier dependency, repair the sequence. If performance is secure only under familiar conditions, add delay or transfer before broadening the conclusion.

Finally, inspect what happens after the decision. Did teaching change? Did the learner improve? Did later performance confirm the inference? Did the curriculum make subsequent learning easier to build? Summative assessment becomes trustworthy when its representations remain correctable by the learner’s return.

61. Prior knowledge

For prior knowledge, summative assessment begins by stating the educational claim and the evidence that could reasonably support it. Ask what capability is being sampled, what conditions shape performance and what decision will follow. Avoid collecting information simply because it is easy to count or administratively familiar.

In the fictional eduKateSG classroom, Adrian compares the intended learning with the actual demand of the task. Jo records important support conditions and inspects patterns rather than treating one response as a complete portrait. When evidence is ambiguous, the next move is another discriminating task or a narrower claim, not greater certainty in the wording.

The action should remain proportionate. If the evidence reveals a prerequisite gap, repair it. If the task samples too little, gather more evidence. If the curriculum introduces a later demand without an earlier dependency, repair the sequence. If performance is secure only under familiar conditions, add delay or transfer before broadening the conclusion.

Finally, inspect what happens after the decision. Did teaching change? Did the learner improve? Did later performance confirm the inference? Did the curriculum make subsequent learning easier to build? Summative assessment becomes trustworthy when its representations remain correctable by the learner’s return.

62. Cognitive load

For cognitive load, summative assessment begins by stating the educational claim and the evidence that could reasonably support it. Ask what capability is being sampled, what conditions shape performance and what decision will follow. Avoid collecting information simply because it is easy to count or administratively familiar.

In the fictional eduKateSG classroom, Adrian compares the intended learning with the actual demand of the task. Jo records important support conditions and inspects patterns rather than treating one response as a complete portrait. When evidence is ambiguous, the next move is another discriminating task or a narrower claim, not greater certainty in the wording.

The action should remain proportionate. If the evidence reveals a prerequisite gap, repair it. If the task samples too little, gather more evidence. If the curriculum introduces a later demand without an earlier dependency, repair the sequence. If performance is secure only under familiar conditions, add delay or transfer before broadening the conclusion.

Finally, inspect what happens after the decision. Did teaching change? Did the learner improve? Did later performance confirm the inference? Did the curriculum make subsequent learning easier to build? Summative assessment becomes trustworthy when its representations remain correctable by the learner’s return.

63. Learning environment

For learning environment, summative assessment begins by stating the educational claim and the evidence that could reasonably support it. Ask what capability is being sampled, what conditions shape performance and what decision will follow. Avoid collecting information simply because it is easy to count or administratively familiar.

In the fictional eduKateSG classroom, Adrian compares the intended learning with the actual demand of the task. Jo records important support conditions and inspects patterns rather than treating one response as a complete portrait. When evidence is ambiguous, the next move is another discriminating task or a narrower claim, not greater certainty in the wording.

The action should remain proportionate. If the evidence reveals a prerequisite gap, repair it. If the task samples too little, gather more evidence. If the curriculum introduces a later demand without an earlier dependency, repair the sequence. If performance is secure only under familiar conditions, add delay or transfer before broadening the conclusion.

Finally, inspect what happens after the decision. Did teaching change? Did the learner improve? Did later performance confirm the inference? Did the curriculum make subsequent learning easier to build? Summative assessment becomes trustworthy when its representations remain correctable by the learner’s return.

64. Productive struggle

For productive struggle, summative assessment begins by stating the educational claim and the evidence that could reasonably support it. Ask what capability is being sampled, what conditions shape performance and what decision will follow. Avoid collecting information simply because it is easy to count or administratively familiar.

In the fictional eduKateSG classroom, Adrian compares the intended learning with the actual demand of the task. Jo records important support conditions and inspects patterns rather than treating one response as a complete portrait. When evidence is ambiguous, the next move is another discriminating task or a narrower claim, not greater certainty in the wording.

The action should remain proportionate. If the evidence reveals a prerequisite gap, repair it. If the task samples too little, gather more evidence. If the curriculum introduces a later demand without an earlier dependency, repair the sequence. If performance is secure only under familiar conditions, add delay or transfer before broadening the conclusion.

Finally, inspect what happens after the decision. Did teaching change? Did the learner improve? Did later performance confirm the inference? Did the curriculum make subsequent learning easier to build? Summative assessment becomes trustworthy when its representations remain correctable by the learner’s return.

65. Academic confidence

For academic confidence, summative assessment begins by stating the educational claim and the evidence that could reasonably support it. Ask what capability is being sampled, what conditions shape performance and what decision will follow. Avoid collecting information simply because it is easy to count or administratively familiar.

In the fictional eduKateSG classroom, Adrian compares the intended learning with the actual demand of the task. Jo records important support conditions and inspects patterns rather than treating one response as a complete portrait. When evidence is ambiguous, the next move is another discriminating task or a narrower claim, not greater certainty in the wording.

The action should remain proportionate. If the evidence reveals a prerequisite gap, repair it. If the task samples too little, gather more evidence. If the curriculum introduces a later demand without an earlier dependency, repair the sequence. If performance is secure only under familiar conditions, add delay or transfer before broadening the conclusion.

Finally, inspect what happens after the decision. Did teaching change? Did the learner improve? Did later performance confirm the inference? Did the curriculum make subsequent learning easier to build? Summative assessment becomes trustworthy when its representations remain correctable by the learner’s return.

66. Help-seeking

For help-seeking, summative assessment begins by stating the educational claim and the evidence that could reasonably support it. Ask what capability is being sampled, what conditions shape performance and what decision will follow. Avoid collecting information simply because it is easy to count or administratively familiar.

In the fictional eduKateSG classroom, Adrian compares the intended learning with the actual demand of the task. Jo records important support conditions and inspects patterns rather than treating one response as a complete portrait. When evidence is ambiguous, the next move is another discriminating task or a narrower claim, not greater certainty in the wording.

The action should remain proportionate. If the evidence reveals a prerequisite gap, repair it. If the task samples too little, gather more evidence. If the curriculum introduces a later demand without an earlier dependency, repair the sequence. If performance is secure only under familiar conditions, add delay or transfer before broadening the conclusion.

Finally, inspect what happens after the decision. Did teaching change? Did the learner improve? Did later performance confirm the inference? Did the curriculum make subsequent learning easier to build? Summative assessment becomes trustworthy when its representations remain correctable by the learner’s return.

67. Student independence

For student independence, summative assessment begins by stating the educational claim and the evidence that could reasonably support it. Ask what capability is being sampled, what conditions shape performance and what decision will follow. Avoid collecting information simply because it is easy to count or administratively familiar.

In the fictional eduKateSG classroom, Adrian compares the intended learning with the actual demand of the task. Jo records important support conditions and inspects patterns rather than treating one response as a complete portrait. When evidence is ambiguous, the next move is another discriminating task or a narrower claim, not greater certainty in the wording.

The action should remain proportionate. If the evidence reveals a prerequisite gap, repair it. If the task samples too little, gather more evidence. If the curriculum introduces a later demand without an earlier dependency, repair the sequence. If performance is secure only under familiar conditions, add delay or transfer before broadening the conclusion.

Finally, inspect what happens after the decision. Did teaching change? Did the learner improve? Did later performance confirm the inference? Did the curriculum make subsequent learning easier to build? Summative assessment becomes trustworthy when its representations remain correctable by the learner’s return.

68. Motivation

For motivation, summative assessment begins by stating the educational claim and the evidence that could reasonably support it. Ask what capability is being sampled, what conditions shape performance and what decision will follow. Avoid collecting information simply because it is easy to count or administratively familiar.

In the fictional eduKateSG classroom, Adrian compares the intended learning with the actual demand of the task. Jo records important support conditions and inspects patterns rather than treating one response as a complete portrait. When evidence is ambiguous, the next move is another discriminating task or a narrower claim, not greater certainty in the wording.

The action should remain proportionate. If the evidence reveals a prerequisite gap, repair it. If the task samples too little, gather more evidence. If the curriculum introduces a later demand without an earlier dependency, repair the sequence. If performance is secure only under familiar conditions, add delay or transfer before broadening the conclusion.

Finally, inspect what happens after the decision. Did teaching change? Did the learner improve? Did later performance confirm the inference? Did the curriculum make subsequent learning easier to build? Summative assessment becomes trustworthy when its representations remain correctable by the learner’s return.

69. Student choice

For student choice, summative assessment begins by stating the educational claim and the evidence that could reasonably support it. Ask what capability is being sampled, what conditions shape performance and what decision will follow. Avoid collecting information simply because it is easy to count or administratively familiar.

In the fictional eduKateSG classroom, Adrian compares the intended learning with the actual demand of the task. Jo records important support conditions and inspects patterns rather than treating one response as a complete portrait. When evidence is ambiguous, the next move is another discriminating task or a narrower claim, not greater certainty in the wording.

The action should remain proportionate. If the evidence reveals a prerequisite gap, repair it. If the task samples too little, gather more evidence. If the curriculum introduces a later demand without an earlier dependency, repair the sequence. If performance is secure only under familiar conditions, add delay or transfer before broadening the conclusion.

Finally, inspect what happens after the decision. Did teaching change? Did the learner improve? Did later performance confirm the inference? Did the curriculum make subsequent learning easier to build? Summative assessment becomes trustworthy when its representations remain correctable by the learner’s return.

70. High expectations

For high expectations, summative assessment begins by stating the educational claim and the evidence that could reasonably support it. Ask what capability is being sampled, what conditions shape performance and what decision will follow. Avoid collecting information simply because it is easy to count or administratively familiar.

In the fictional eduKateSG classroom, Adrian compares the intended learning with the actual demand of the task. Jo records important support conditions and inspects patterns rather than treating one response as a complete portrait. When evidence is ambiguous, the next move is another discriminating task or a narrower claim, not greater certainty in the wording.

The action should remain proportionate. If the evidence reveals a prerequisite gap, repair it. If the task samples too little, gather more evidence. If the curriculum introduces a later demand without an earlier dependency, repair the sequence. If performance is secure only under familiar conditions, add delay or transfer before broadening the conclusion.

Finally, inspect what happens after the decision. Did teaching change? Did the learner improve? Did later performance confirm the inference? Did the curriculum make subsequent learning easier to build? Summative assessment becomes trustworthy when its representations remain correctable by the learner’s return.

71. Classroom routines

For classroom routines, summative assessment begins by stating the educational claim and the evidence that could reasonably support it. Ask what capability is being sampled, what conditions shape performance and what decision will follow. Avoid collecting information simply because it is easy to count or administratively familiar.

In the fictional eduKateSG classroom, Adrian compares the intended learning with the actual demand of the task. Jo records important support conditions and inspects patterns rather than treating one response as a complete portrait. When evidence is ambiguous, the next move is another discriminating task or a narrower claim, not greater certainty in the wording.

The action should remain proportionate. If the evidence reveals a prerequisite gap, repair it. If the task samples too little, gather more evidence. If the curriculum introduces a later demand without an earlier dependency, repair the sequence. If performance is secure only under familiar conditions, add delay or transfer before broadening the conclusion.

Finally, inspect what happens after the decision. Did teaching change? Did the learner improve? Did later performance confirm the inference? Did the curriculum make subsequent learning easier to build? Summative assessment becomes trustworthy when its representations remain correctable by the learner’s return.

72. Teacher modelling

For teacher modelling, summative assessment begins by stating the educational claim and the evidence that could reasonably support it. Ask what capability is being sampled, what conditions shape performance and what decision will follow. Avoid collecting information simply because it is easy to count or administratively familiar.

In the fictional eduKateSG classroom, Adrian compares the intended learning with the actual demand of the task. Jo records important support conditions and inspects patterns rather than treating one response as a complete portrait. When evidence is ambiguous, the next move is another discriminating task or a narrower claim, not greater certainty in the wording.

The action should remain proportionate. If the evidence reveals a prerequisite gap, repair it. If the task samples too little, gather more evidence. If the curriculum introduces a later demand without an earlier dependency, repair the sequence. If performance is secure only under familiar conditions, add delay or transfer before broadening the conclusion.

Finally, inspect what happens after the decision. Did teaching change? Did the learner improve? Did later performance confirm the inference? Did the curriculum make subsequent learning easier to build? Summative assessment becomes trustworthy when its representations remain correctable by the learner’s return.

73. Scaffolding

For scaffolding, summative assessment begins by stating the educational claim and the evidence that could reasonably support it. Ask what capability is being sampled, what conditions shape performance and what decision will follow. Avoid collecting information simply because it is easy to count or administratively familiar.

In the fictional eduKateSG classroom, Adrian compares the intended learning with the actual demand of the task. Jo records important support conditions and inspects patterns rather than treating one response as a complete portrait. When evidence is ambiguous, the next move is another discriminating task or a narrower claim, not greater certainty in the wording.

The action should remain proportionate. If the evidence reveals a prerequisite gap, repair it. If the task samples too little, gather more evidence. If the curriculum introduces a later demand without an earlier dependency, repair the sequence. If performance is secure only under familiar conditions, add delay or transfer before broadening the conclusion.

Finally, inspect what happens after the decision. Did teaching change? Did the learner improve? Did later performance confirm the inference? Did the curriculum make subsequent learning easier to build? Summative assessment becomes trustworthy when its representations remain correctable by the learner’s return.

74. Classroom discussion

For classroom discussion, summative assessment begins by stating the educational claim and the evidence that could reasonably support it. Ask what capability is being sampled, what conditions shape performance and what decision will follow. Avoid collecting information simply because it is easy to count or administratively familiar.

In the fictional eduKateSG classroom, Adrian compares the intended learning with the actual demand of the task. Jo records important support conditions and inspects patterns rather than treating one response as a complete portrait. When evidence is ambiguous, the next move is another discriminating task or a narrower claim, not greater certainty in the wording.

The action should remain proportionate. If the evidence reveals a prerequisite gap, repair it. If the task samples too little, gather more evidence. If the curriculum introduces a later demand without an earlier dependency, repair the sequence. If performance is secure only under familiar conditions, add delay or transfer before broadening the conclusion.

Finally, inspect what happens after the decision. Did teaching change? Did the learner improve? Did later performance confirm the inference? Did the curriculum make subsequent learning easier to build? Summative assessment becomes trustworthy when its representations remain correctable by the learner’s return.

75. Collaborative learning

For collaborative learning, summative assessment begins by stating the educational claim and the evidence that could reasonably support it. Ask what capability is being sampled, what conditions shape performance and what decision will follow. Avoid collecting information simply because it is easy to count or administratively familiar.

In the fictional eduKateSG classroom, Adrian compares the intended learning with the actual demand of the task. Jo records important support conditions and inspects patterns rather than treating one response as a complete portrait. When evidence is ambiguous, the next move is another discriminating task or a narrower claim, not greater certainty in the wording.

The action should remain proportionate. If the evidence reveals a prerequisite gap, repair it. If the task samples too little, gather more evidence. If the curriculum introduces a later demand without an earlier dependency, repair the sequence. If performance is secure only under familiar conditions, add delay or transfer before broadening the conclusion.

Finally, inspect what happens after the decision. Did teaching change? Did the learner improve? Did later performance confirm the inference? Did the curriculum make subsequent learning easier to build? Summative assessment becomes trustworthy when its representations remain correctable by the learner’s return.

76. Peer feedback

For peer feedback, summative assessment begins by stating the educational claim and the evidence that could reasonably support it. Ask what capability is being sampled, what conditions shape performance and what decision will follow. Avoid collecting information simply because it is easy to count or administratively familiar.

In the fictional eduKateSG classroom, Adrian compares the intended learning with the actual demand of the task. Jo records important support conditions and inspects patterns rather than treating one response as a complete portrait. When evidence is ambiguous, the next move is another discriminating task or a narrower claim, not greater certainty in the wording.

The action should remain proportionate. If the evidence reveals a prerequisite gap, repair it. If the task samples too little, gather more evidence. If the curriculum introduces a later demand without an earlier dependency, repair the sequence. If performance is secure only under familiar conditions, add delay or transfer before broadening the conclusion.

Finally, inspect what happens after the decision. Did teaching change? Did the learner improve? Did later performance confirm the inference? Did the curriculum make subsequent learning easier to build? Summative assessment becomes trustworthy when its representations remain correctable by the learner’s return.

77. Homework

For homework, summative assessment begins by stating the educational claim and the evidence that could reasonably support it. Ask what capability is being sampled, what conditions shape performance and what decision will follow. Avoid collecting information simply because it is easy to count or administratively familiar.

In the fictional eduKateSG classroom, Adrian compares the intended learning with the actual demand of the task. Jo records important support conditions and inspects patterns rather than treating one response as a complete portrait. When evidence is ambiguous, the next move is another discriminating task or a narrower claim, not greater certainty in the wording.

The action should remain proportionate. If the evidence reveals a prerequisite gap, repair it. If the task samples too little, gather more evidence. If the curriculum introduces a later demand without an earlier dependency, repair the sequence. If performance is secure only under familiar conditions, add delay or transfer before broadening the conclusion.

Finally, inspect what happens after the decision. Did teaching change? Did the learner improve? Did later performance confirm the inference? Did the curriculum make subsequent learning easier to build? Summative assessment becomes trustworthy when its representations remain correctable by the learner’s return.

78. Study skills

For study skills, summative assessment begins by stating the educational claim and the evidence that could reasonably support it. Ask what capability is being sampled, what conditions shape performance and what decision will follow. Avoid collecting information simply because it is easy to count or administratively familiar.

In the fictional eduKateSG classroom, Adrian compares the intended learning with the actual demand of the task. Jo records important support conditions and inspects patterns rather than treating one response as a complete portrait. When evidence is ambiguous, the next move is another discriminating task or a narrower claim, not greater certainty in the wording.

The action should remain proportionate. If the evidence reveals a prerequisite gap, repair it. If the task samples too little, gather more evidence. If the curriculum introduces a later demand without an earlier dependency, repair the sequence. If performance is secure only under familiar conditions, add delay or transfer before broadening the conclusion.

Finally, inspect what happens after the decision. Did teaching change? Did the learner improve? Did later performance confirm the inference? Did the curriculum make subsequent learning easier to build? Summative assessment becomes trustworthy when its representations remain correctable by the learner’s return.

79. Mathematics

For mathematics, summative assessment begins by stating the educational claim and the evidence that could reasonably support it. Ask what capability is being sampled, what conditions shape performance and what decision will follow. Avoid collecting information simply because it is easy to count or administratively familiar.

In the fictional eduKateSG classroom, Adrian compares the intended learning with the actual demand of the task. Jo records important support conditions and inspects patterns rather than treating one response as a complete portrait. When evidence is ambiguous, the next move is another discriminating task or a narrower claim, not greater certainty in the wording.

The action should remain proportionate. If the evidence reveals a prerequisite gap, repair it. If the task samples too little, gather more evidence. If the curriculum introduces a later demand without an earlier dependency, repair the sequence. If performance is secure only under familiar conditions, add delay or transfer before broadening the conclusion.

Finally, inspect what happens after the decision. Did teaching change? Did the learner improve? Did later performance confirm the inference? Did the curriculum make subsequent learning easier to build? Summative assessment becomes trustworthy when its representations remain correctable by the learner’s return.

80. Science

For science, summative assessment begins by stating the educational claim and the evidence that could reasonably support it. Ask what capability is being sampled, what conditions shape performance and what decision will follow. Avoid collecting information simply because it is easy to count or administratively familiar.

In the fictional eduKateSG classroom, Adrian compares the intended learning with the actual demand of the task. Jo records important support conditions and inspects patterns rather than treating one response as a complete portrait. When evidence is ambiguous, the next move is another discriminating task or a narrower claim, not greater certainty in the wording.

The action should remain proportionate. If the evidence reveals a prerequisite gap, repair it. If the task samples too little, gather more evidence. If the curriculum introduces a later demand without an earlier dependency, repair the sequence. If performance is secure only under familiar conditions, add delay or transfer before broadening the conclusion.

Finally, inspect what happens after the decision. Did teaching change? Did the learner improve? Did later performance confirm the inference? Did the curriculum make subsequent learning easier to build? Summative assessment becomes trustworthy when its representations remain correctable by the learner’s return.

81. History

For history, summative assessment begins by stating the educational claim and the evidence that could reasonably support it. Ask what capability is being sampled, what conditions shape performance and what decision will follow. Avoid collecting information simply because it is easy to count or administratively familiar.

In the fictional eduKateSG classroom, Adrian compares the intended learning with the actual demand of the task. Jo records important support conditions and inspects patterns rather than treating one response as a complete portrait. When evidence is ambiguous, the next move is another discriminating task or a narrower claim, not greater certainty in the wording.

The action should remain proportionate. If the evidence reveals a prerequisite gap, repair it. If the task samples too little, gather more evidence. If the curriculum introduces a later demand without an earlier dependency, repair the sequence. If performance is secure only under familiar conditions, add delay or transfer before broadening the conclusion.

Finally, inspect what happens after the decision. Did teaching change? Did the learner improve? Did later performance confirm the inference? Did the curriculum make subsequent learning easier to build? Summative assessment becomes trustworthy when its representations remain correctable by the learner’s return.

82. Geography

For geography, summative assessment begins by stating the educational claim and the evidence that could reasonably support it. Ask what capability is being sampled, what conditions shape performance and what decision will follow. Avoid collecting information simply because it is easy to count or administratively familiar.

In the fictional eduKateSG classroom, Adrian compares the intended learning with the actual demand of the task. Jo records important support conditions and inspects patterns rather than treating one response as a complete portrait. When evidence is ambiguous, the next move is another discriminating task or a narrower claim, not greater certainty in the wording.

The action should remain proportionate. If the evidence reveals a prerequisite gap, repair it. If the task samples too little, gather more evidence. If the curriculum introduces a later demand without an earlier dependency, repair the sequence. If performance is secure only under familiar conditions, add delay or transfer before broadening the conclusion.

Finally, inspect what happens after the decision. Did teaching change? Did the learner improve? Did later performance confirm the inference? Did the curriculum make subsequent learning easier to build? Summative assessment becomes trustworthy when its representations remain correctable by the learner’s return.

83. Reading

For reading, summative assessment begins by stating the educational claim and the evidence that could reasonably support it. Ask what capability is being sampled, what conditions shape performance and what decision will follow. Avoid collecting information simply because it is easy to count or administratively familiar.

In the fictional eduKateSG classroom, Adrian compares the intended learning with the actual demand of the task. Jo records important support conditions and inspects patterns rather than treating one response as a complete portrait. When evidence is ambiguous, the next move is another discriminating task or a narrower claim, not greater certainty in the wording.

The action should remain proportionate. If the evidence reveals a prerequisite gap, repair it. If the task samples too little, gather more evidence. If the curriculum introduces a later demand without an earlier dependency, repair the sequence. If performance is secure only under familiar conditions, add delay or transfer before broadening the conclusion.

Finally, inspect what happens after the decision. Did teaching change? Did the learner improve? Did later performance confirm the inference? Did the curriculum make subsequent learning easier to build? Summative assessment becomes trustworthy when its representations remain correctable by the learner’s return.

84. Vocabulary

For vocabulary, summative assessment begins by stating the educational claim and the evidence that could reasonably support it. Ask what capability is being sampled, what conditions shape performance and what decision will follow. Avoid collecting information simply because it is easy to count or administratively familiar.

In the fictional eduKateSG classroom, Adrian compares the intended learning with the actual demand of the task. Jo records important support conditions and inspects patterns rather than treating one response as a complete portrait. When evidence is ambiguous, the next move is another discriminating task or a narrower claim, not greater certainty in the wording.

The action should remain proportionate. If the evidence reveals a prerequisite gap, repair it. If the task samples too little, gather more evidence. If the curriculum introduces a later demand without an earlier dependency, repair the sequence. If performance is secure only under familiar conditions, add delay or transfer before broadening the conclusion.

Finally, inspect what happens after the decision. Did teaching change? Did the learner improve? Did later performance confirm the inference? Did the curriculum make subsequent learning easier to build? Summative assessment becomes trustworthy when its representations remain correctable by the learner’s return.

85. English

For english, summative assessment begins by stating the educational claim and the evidence that could reasonably support it. Ask what capability is being sampled, what conditions shape performance and what decision will follow. Avoid collecting information simply because it is easy to count or administratively familiar.

In the fictional eduKateSG classroom, Adrian compares the intended learning with the actual demand of the task. Jo records important support conditions and inspects patterns rather than treating one response as a complete portrait. When evidence is ambiguous, the next move is another discriminating task or a narrower claim, not greater certainty in the wording.

The action should remain proportionate. If the evidence reveals a prerequisite gap, repair it. If the task samples too little, gather more evidence. If the curriculum introduces a later demand without an earlier dependency, repair the sequence. If performance is secure only under familiar conditions, add delay or transfer before broadening the conclusion.

Finally, inspect what happens after the decision. Did teaching change? Did the learner improve? Did later performance confirm the inference? Did the curriculum make subsequent learning easier to build? Summative assessment becomes trustworthy when its representations remain correctable by the learner’s return.

86. Creative writing

For creative writing, summative assessment begins by stating the educational claim and the evidence that could reasonably support it. Ask what capability is being sampled, what conditions shape performance and what decision will follow. Avoid collecting information simply because it is easy to count or administratively familiar.

In the fictional eduKateSG classroom, Adrian compares the intended learning with the actual demand of the task. Jo records important support conditions and inspects patterns rather than treating one response as a complete portrait. When evidence is ambiguous, the next move is another discriminating task or a narrower claim, not greater certainty in the wording.

The action should remain proportionate. If the evidence reveals a prerequisite gap, repair it. If the task samples too little, gather more evidence. If the curriculum introduces a later demand without an earlier dependency, repair the sequence. If performance is secure only under familiar conditions, add delay or transfer before broadening the conclusion.

Finally, inspect what happens after the decision. Did teaching change? Did the learner improve? Did later performance confirm the inference? Did the curriculum make subsequent learning easier to build? Summative assessment becomes trustworthy when its representations remain correctable by the learner’s return.

87. Oral communication

For oral communication, summative assessment begins by stating the educational claim and the evidence that could reasonably support it. Ask what capability is being sampled, what conditions shape performance and what decision will follow. Avoid collecting information simply because it is easy to count or administratively familiar.

In the fictional eduKateSG classroom, Adrian compares the intended learning with the actual demand of the task. Jo records important support conditions and inspects patterns rather than treating one response as a complete portrait. When evidence is ambiguous, the next move is another discriminating task or a narrower claim, not greater certainty in the wording.

The action should remain proportionate. If the evidence reveals a prerequisite gap, repair it. If the task samples too little, gather more evidence. If the curriculum introduces a later demand without an earlier dependency, repair the sequence. If performance is secure only under familiar conditions, add delay or transfer before broadening the conclusion.

Finally, inspect what happens after the decision. Did teaching change? Did the learner improve? Did later performance confirm the inference? Did the curriculum make subsequent learning easier to build? Summative assessment becomes trustworthy when its representations remain correctable by the learner’s return.

88. Research

For research, summative assessment begins by stating the educational claim and the evidence that could reasonably support it. Ask what capability is being sampled, what conditions shape performance and what decision will follow. Avoid collecting information simply because it is easy to count or administratively familiar.

In the fictional eduKateSG classroom, Adrian compares the intended learning with the actual demand of the task. Jo records important support conditions and inspects patterns rather than treating one response as a complete portrait. When evidence is ambiguous, the next move is another discriminating task or a narrower claim, not greater certainty in the wording.

The action should remain proportionate. If the evidence reveals a prerequisite gap, repair it. If the task samples too little, gather more evidence. If the curriculum introduces a later demand without an earlier dependency, repair the sequence. If performance is secure only under familiar conditions, add delay or transfer before broadening the conclusion.

Finally, inspect what happens after the decision. Did teaching change? Did the learner improve? Did later performance confirm the inference? Did the curriculum make subsequent learning easier to build? Summative assessment becomes trustworthy when its representations remain correctable by the learner’s return.

89. Projects

For projects, summative assessment begins by stating the educational claim and the evidence that could reasonably support it. Ask what capability is being sampled, what conditions shape performance and what decision will follow. Avoid collecting information simply because it is easy to count or administratively familiar.

In the fictional eduKateSG classroom, Adrian compares the intended learning with the actual demand of the task. Jo records important support conditions and inspects patterns rather than treating one response as a complete portrait. When evidence is ambiguous, the next move is another discriminating task or a narrower claim, not greater certainty in the wording.

The action should remain proportionate. If the evidence reveals a prerequisite gap, repair it. If the task samples too little, gather more evidence. If the curriculum introduces a later demand without an earlier dependency, repair the sequence. If performance is secure only under familiar conditions, add delay or transfer before broadening the conclusion.

Finally, inspect what happens after the decision. Did teaching change? Did the learner improve? Did later performance confirm the inference? Did the curriculum make subsequent learning easier to build? Summative assessment becomes trustworthy when its representations remain correctable by the learner’s return.

90. Interdisciplinary learning

For interdisciplinary learning, summative assessment begins by stating the educational claim and the evidence that could reasonably support it. Ask what capability is being sampled, what conditions shape performance and what decision will follow. Avoid collecting information simply because it is easy to count or administratively familiar.

In the fictional eduKateSG classroom, Adrian compares the intended learning with the actual demand of the task. Jo records important support conditions and inspects patterns rather than treating one response as a complete portrait. When evidence is ambiguous, the next move is another discriminating task or a narrower claim, not greater certainty in the wording.

The action should remain proportionate. If the evidence reveals a prerequisite gap, repair it. If the task samples too little, gather more evidence. If the curriculum introduces a later demand without an earlier dependency, repair the sequence. If performance is secure only under familiar conditions, add delay or transfer before broadening the conclusion.

Finally, inspect what happens after the decision. Did teaching change? Did the learner improve? Did later performance confirm the inference? Did the curriculum make subsequent learning easier to build? Summative assessment becomes trustworthy when its representations remain correctable by the learner’s return.

91. Primary learners

For primary learners, summative assessment begins by stating the educational claim and the evidence that could reasonably support it. Ask what capability is being sampled, what conditions shape performance and what decision will follow. Avoid collecting information simply because it is easy to count or administratively familiar.

In the fictional eduKateSG classroom, Adrian compares the intended learning with the actual demand of the task. Jo records important support conditions and inspects patterns rather than treating one response as a complete portrait. When evidence is ambiguous, the next move is another discriminating task or a narrower claim, not greater certainty in the wording.

The action should remain proportionate. If the evidence reveals a prerequisite gap, repair it. If the task samples too little, gather more evidence. If the curriculum introduces a later demand without an earlier dependency, repair the sequence. If performance is secure only under familiar conditions, add delay or transfer before broadening the conclusion.

Finally, inspect what happens after the decision. Did teaching change? Did the learner improve? Did later performance confirm the inference? Did the curriculum make subsequent learning easier to build? Summative assessment becomes trustworthy when its representations remain correctable by the learner’s return.

92. Secondary learners

For secondary learners, summative assessment begins by stating the educational claim and the evidence that could reasonably support it. Ask what capability is being sampled, what conditions shape performance and what decision will follow. Avoid collecting information simply because it is easy to count or administratively familiar.

In the fictional eduKateSG classroom, Adrian compares the intended learning with the actual demand of the task. Jo records important support conditions and inspects patterns rather than treating one response as a complete portrait. When evidence is ambiguous, the next move is another discriminating task or a narrower claim, not greater certainty in the wording.

The action should remain proportionate. If the evidence reveals a prerequisite gap, repair it. If the task samples too little, gather more evidence. If the curriculum introduces a later demand without an earlier dependency, repair the sequence. If performance is secure only under familiar conditions, add delay or transfer before broadening the conclusion.

Finally, inspect what happens after the decision. Did teaching change? Did the learner improve? Did later performance confirm the inference? Did the curriculum make subsequent learning easier to build? Summative assessment becomes trustworthy when its representations remain correctable by the learner’s return.

93. Advanced learners

For advanced learners, summative assessment begins by stating the educational claim and the evidence that could reasonably support it. Ask what capability is being sampled, what conditions shape performance and what decision will follow. Avoid collecting information simply because it is easy to count or administratively familiar.

In the fictional eduKateSG classroom, Adrian compares the intended learning with the actual demand of the task. Jo records important support conditions and inspects patterns rather than treating one response as a complete portrait. When evidence is ambiguous, the next move is another discriminating task or a narrower claim, not greater certainty in the wording.

The action should remain proportionate. If the evidence reveals a prerequisite gap, repair it. If the task samples too little, gather more evidence. If the curriculum introduces a later demand without an earlier dependency, repair the sequence. If performance is secure only under familiar conditions, add delay or transfer before broadening the conclusion.

Finally, inspect what happens after the decision. Did teaching change? Did the learner improve? Did later performance confirm the inference? Did the curriculum make subsequent learning easier to build? Summative assessment becomes trustworthy when its representations remain correctable by the learner’s return.

94. Multilingual learners

For multilingual learners, summative assessment begins by stating the educational claim and the evidence that could reasonably support it. Ask what capability is being sampled, what conditions shape performance and what decision will follow. Avoid collecting information simply because it is easy to count or administratively familiar.

In the fictional eduKateSG classroom, Adrian compares the intended learning with the actual demand of the task. Jo records important support conditions and inspects patterns rather than treating one response as a complete portrait. When evidence is ambiguous, the next move is another discriminating task or a narrower claim, not greater certainty in the wording.

The action should remain proportionate. If the evidence reveals a prerequisite gap, repair it. If the task samples too little, gather more evidence. If the curriculum introduces a later demand without an earlier dependency, repair the sequence. If performance is secure only under familiar conditions, add delay or transfer before broadening the conclusion.

Finally, inspect what happens after the decision. Did teaching change? Did the learner improve? Did later performance confirm the inference? Did the curriculum make subsequent learning easier to build? Summative assessment becomes trustworthy when its representations remain correctable by the learner’s return.

95. Parents

For parents, summative assessment begins by stating the educational claim and the evidence that could reasonably support it. Ask what capability is being sampled, what conditions shape performance and what decision will follow. Avoid collecting information simply because it is easy to count or administratively familiar.

In the fictional eduKateSG classroom, Adrian compares the intended learning with the actual demand of the task. Jo records important support conditions and inspects patterns rather than treating one response as a complete portrait. When evidence is ambiguous, the next move is another discriminating task or a narrower claim, not greater certainty in the wording.

The action should remain proportionate. If the evidence reveals a prerequisite gap, repair it. If the task samples too little, gather more evidence. If the curriculum introduces a later demand without an earlier dependency, repair the sequence. If performance is secure only under familiar conditions, add delay or transfer before broadening the conclusion.

Finally, inspect what happens after the decision. Did teaching change? Did the learner improve? Did later performance confirm the inference? Did the curriculum make subsequent learning easier to build? Summative assessment becomes trustworthy when its representations remain correctable by the learner’s return.

96. Tutors

For tutors, summative assessment begins by stating the educational claim and the evidence that could reasonably support it. Ask what capability is being sampled, what conditions shape performance and what decision will follow. Avoid collecting information simply because it is easy to count or administratively familiar.

In the fictional eduKateSG classroom, Adrian compares the intended learning with the actual demand of the task. Jo records important support conditions and inspects patterns rather than treating one response as a complete portrait. When evidence is ambiguous, the next move is another discriminating task or a narrower claim, not greater certainty in the wording.

The action should remain proportionate. If the evidence reveals a prerequisite gap, repair it. If the task samples too little, gather more evidence. If the curriculum introduces a later demand without an earlier dependency, repair the sequence. If performance is secure only under familiar conditions, add delay or transfer before broadening the conclusion.

Finally, inspect what happens after the decision. Did teaching change? Did the learner improve? Did later performance confirm the inference? Did the curriculum make subsequent learning easier to build? Summative assessment becomes trustworthy when its representations remain correctable by the learner’s return.

97. Teacher teams

For teacher teams, summative assessment begins by stating the educational claim and the evidence that could reasonably support it. Ask what capability is being sampled, what conditions shape performance and what decision will follow. Avoid collecting information simply because it is easy to count or administratively familiar.

In the fictional eduKateSG classroom, Adrian compares the intended learning with the actual demand of the task. Jo records important support conditions and inspects patterns rather than treating one response as a complete portrait. When evidence is ambiguous, the next move is another discriminating task or a narrower claim, not greater certainty in the wording.

The action should remain proportionate. If the evidence reveals a prerequisite gap, repair it. If the task samples too little, gather more evidence. If the curriculum introduces a later demand without an earlier dependency, repair the sequence. If performance is secure only under familiar conditions, add delay or transfer before broadening the conclusion.

Finally, inspect what happens after the decision. Did teaching change? Did the learner improve? Did later performance confirm the inference? Did the curriculum make subsequent learning easier to build? Summative assessment becomes trustworthy when its representations remain correctable by the learner’s return.

98. Curriculum teams

For curriculum teams, summative assessment begins by stating the educational claim and the evidence that could reasonably support it. Ask what capability is being sampled, what conditions shape performance and what decision will follow. Avoid collecting information simply because it is easy to count or administratively familiar.

In the fictional eduKateSG classroom, Adrian compares the intended learning with the actual demand of the task. Jo records important support conditions and inspects patterns rather than treating one response as a complete portrait. When evidence is ambiguous, the next move is another discriminating task or a narrower claim, not greater certainty in the wording.

The action should remain proportionate. If the evidence reveals a prerequisite gap, repair it. If the task samples too little, gather more evidence. If the curriculum introduces a later demand without an earlier dependency, repair the sequence. If performance is secure only under familiar conditions, add delay or transfer before broadening the conclusion.

Finally, inspect what happens after the decision. Did teaching change? Did the learner improve? Did later performance confirm the inference? Did the curriculum make subsequent learning easier to build? Summative assessment becomes trustworthy when its representations remain correctable by the learner’s return.

99. School leaders

For school leaders, summative assessment begins by stating the educational claim and the evidence that could reasonably support it. Ask what capability is being sampled, what conditions shape performance and what decision will follow. Avoid collecting information simply because it is easy to count or administratively familiar.

In the fictional eduKateSG classroom, Adrian compares the intended learning with the actual demand of the task. Jo records important support conditions and inspects patterns rather than treating one response as a complete portrait. When evidence is ambiguous, the next move is another discriminating task or a narrower claim, not greater certainty in the wording.

The action should remain proportionate. If the evidence reveals a prerequisite gap, repair it. If the task samples too little, gather more evidence. If the curriculum introduces a later demand without an earlier dependency, repair the sequence. If performance is secure only under familiar conditions, add delay or transfer before broadening the conclusion.

Finally, inspect what happens after the decision. Did teaching change? Did the learner improve? Did later performance confirm the inference? Did the curriculum make subsequent learning easier to build? Summative assessment becomes trustworthy when its representations remain correctable by the learner’s return.

100. Lesson sequence

For lesson sequence, summative assessment begins by stating the educational claim and the evidence that could reasonably support it. Ask what capability is being sampled, what conditions shape performance and what decision will follow. Avoid collecting information simply because it is easy to count or administratively familiar.

In the fictional eduKateSG classroom, Adrian compares the intended learning with the actual demand of the task. Jo records important support conditions and inspects patterns rather than treating one response as a complete portrait. When evidence is ambiguous, the next move is another discriminating task or a narrower claim, not greater certainty in the wording.

The action should remain proportionate. If the evidence reveals a prerequisite gap, repair it. If the task samples too little, gather more evidence. If the curriculum introduces a later demand without an earlier dependency, repair the sequence. If performance is secure only under familiar conditions, add delay or transfer before broadening the conclusion.

Finally, inspect what happens after the decision. Did teaching change? Did the learner improve? Did later performance confirm the inference? Did the curriculum make subsequent learning easier to build? Summative assessment becomes trustworthy when its representations remain correctable by the learner’s return.

101. Unit sequence

For unit sequence, summative assessment begins by stating the educational claim and the evidence that could reasonably support it. Ask what capability is being sampled, what conditions shape performance and what decision will follow. Avoid collecting information simply because it is easy to count or administratively familiar.

In the fictional eduKateSG classroom, Adrian compares the intended learning with the actual demand of the task. Jo records important support conditions and inspects patterns rather than treating one response as a complete portrait. When evidence is ambiguous, the next move is another discriminating task or a narrower claim, not greater certainty in the wording.

The action should remain proportionate. If the evidence reveals a prerequisite gap, repair it. If the task samples too little, gather more evidence. If the curriculum introduces a later demand without an earlier dependency, repair the sequence. If performance is secure only under familiar conditions, add delay or transfer before broadening the conclusion.

Finally, inspect what happens after the decision. Did teaching change? Did the learner improve? Did later performance confirm the inference? Did the curriculum make subsequent learning easier to build? Summative assessment becomes trustworthy when its representations remain correctable by the learner’s return.

102. Year sequence

For year sequence, summative assessment begins by stating the educational claim and the evidence that could reasonably support it. Ask what capability is being sampled, what conditions shape performance and what decision will follow. Avoid collecting information simply because it is easy to count or administratively familiar.

In the fictional eduKateSG classroom, Adrian compares the intended learning with the actual demand of the task. Jo records important support conditions and inspects patterns rather than treating one response as a complete portrait. When evidence is ambiguous, the next move is another discriminating task or a narrower claim, not greater certainty in the wording.

The action should remain proportionate. If the evidence reveals a prerequisite gap, repair it. If the task samples too little, gather more evidence. If the curriculum introduces a later demand without an earlier dependency, repair the sequence. If performance is secure only under familiar conditions, add delay or transfer before broadening the conclusion.

Finally, inspect what happens after the decision. Did teaching change? Did the learner improve? Did later performance confirm the inference? Did the curriculum make subsequent learning easier to build? Summative assessment becomes trustworthy when its representations remain correctable by the learner’s return.

103. Cumulative review

For cumulative review, summative assessment begins by stating the educational claim and the evidence that could reasonably support it. Ask what capability is being sampled, what conditions shape performance and what decision will follow. Avoid collecting information simply because it is easy to count or administratively familiar.

In the fictional eduKateSG classroom, Adrian compares the intended learning with the actual demand of the task. Jo records important support conditions and inspects patterns rather than treating one response as a complete portrait. When evidence is ambiguous, the next move is another discriminating task or a narrower claim, not greater certainty in the wording.

The action should remain proportionate. If the evidence reveals a prerequisite gap, repair it. If the task samples too little, gather more evidence. If the curriculum introduces a later demand without an earlier dependency, repair the sequence. If performance is secure only under familiar conditions, add delay or transfer before broadening the conclusion.

Finally, inspect what happens after the decision. Did teaching change? Did the learner improve? Did later performance confirm the inference? Did the curriculum make subsequent learning easier to build? Summative assessment becomes trustworthy when its representations remain correctable by the learner’s return.

104. Milestones

For milestones, summative assessment begins by stating the educational claim and the evidence that could reasonably support it. Ask what capability is being sampled, what conditions shape performance and what decision will follow. Avoid collecting information simply because it is easy to count or administratively familiar.

In the fictional eduKateSG classroom, Adrian compares the intended learning with the actual demand of the task. Jo records important support conditions and inspects patterns rather than treating one response as a complete portrait. When evidence is ambiguous, the next move is another discriminating task or a narrower claim, not greater certainty in the wording.

The action should remain proportionate. If the evidence reveals a prerequisite gap, repair it. If the task samples too little, gather more evidence. If the curriculum introduces a later demand without an earlier dependency, repair the sequence. If performance is secure only under familiar conditions, add delay or transfer before broadening the conclusion.

Finally, inspect what happens after the decision. Did teaching change? Did the learner improve? Did later performance confirm the inference? Did the curriculum make subsequent learning easier to build? Summative assessment becomes trustworthy when its representations remain correctable by the learner’s return.

105. Transition

For transition, summative assessment begins by stating the educational claim and the evidence that could reasonably support it. Ask what capability is being sampled, what conditions shape performance and what decision will follow. Avoid collecting information simply because it is easy to count or administratively familiar.

In the fictional eduKateSG classroom, Adrian compares the intended learning with the actual demand of the task. Jo records important support conditions and inspects patterns rather than treating one response as a complete portrait. When evidence is ambiguous, the next move is another discriminating task or a narrower claim, not greater certainty in the wording.

The action should remain proportionate. If the evidence reveals a prerequisite gap, repair it. If the task samples too little, gather more evidence. If the curriculum introduces a later demand without an earlier dependency, repair the sequence. If performance is secure only under familiar conditions, add delay or transfer before broadening the conclusion.

Finally, inspect what happens after the decision. Did teaching change? Did the learner improve? Did later performance confirm the inference? Did the curriculum make subsequent learning easier to build? Summative assessment becomes trustworthy when its representations remain correctable by the learner’s return.

106. Vertical alignment

For vertical alignment, summative assessment begins by stating the educational claim and the evidence that could reasonably support it. Ask what capability is being sampled, what conditions shape performance and what decision will follow. Avoid collecting information simply because it is easy to count or administratively familiar.

In the fictional eduKateSG classroom, Adrian compares the intended learning with the actual demand of the task. Jo records important support conditions and inspects patterns rather than treating one response as a complete portrait. When evidence is ambiguous, the next move is another discriminating task or a narrower claim, not greater certainty in the wording.

The action should remain proportionate. If the evidence reveals a prerequisite gap, repair it. If the task samples too little, gather more evidence. If the curriculum introduces a later demand without an earlier dependency, repair the sequence. If performance is secure only under familiar conditions, add delay or transfer before broadening the conclusion.

Finally, inspect what happens after the decision. Did teaching change? Did the learner improve? Did later performance confirm the inference? Did the curriculum make subsequent learning easier to build? Summative assessment becomes trustworthy when its representations remain correctable by the learner’s return.

107. Horizontal alignment

For horizontal alignment, summative assessment begins by stating the educational claim and the evidence that could reasonably support it. Ask what capability is being sampled, what conditions shape performance and what decision will follow. Avoid collecting information simply because it is easy to count or administratively familiar.

In the fictional eduKateSG classroom, Adrian compares the intended learning with the actual demand of the task. Jo records important support conditions and inspects patterns rather than treating one response as a complete portrait. When evidence is ambiguous, the next move is another discriminating task or a narrower claim, not greater certainty in the wording.

The action should remain proportionate. If the evidence reveals a prerequisite gap, repair it. If the task samples too little, gather more evidence. If the curriculum introduces a later demand without an earlier dependency, repair the sequence. If performance is secure only under familiar conditions, add delay or transfer before broadening the conclusion.

Finally, inspect what happens after the decision. Did teaching change? Did the learner improve? Did later performance confirm the inference? Did the curriculum make subsequent learning easier to build? Summative assessment becomes trustworthy when its representations remain correctable by the learner’s return.

108. Assessment alignment

For assessment alignment, summative assessment begins by stating the educational claim and the evidence that could reasonably support it. Ask what capability is being sampled, what conditions shape performance and what decision will follow. Avoid collecting information simply because it is easy to count or administratively familiar.

In the fictional eduKateSG classroom, Adrian compares the intended learning with the actual demand of the task. Jo records important support conditions and inspects patterns rather than treating one response as a complete portrait. When evidence is ambiguous, the next move is another discriminating task or a narrower claim, not greater certainty in the wording.

The action should remain proportionate. If the evidence reveals a prerequisite gap, repair it. If the task samples too little, gather more evidence. If the curriculum introduces a later demand without an earlier dependency, repair the sequence. If performance is secure only under familiar conditions, add delay or transfer before broadening the conclusion.

Finally, inspect what happens after the decision. Did teaching change? Did the learner improve? Did later performance confirm the inference? Did the curriculum make subsequent learning easier to build? Summative assessment becomes trustworthy when its representations remain correctable by the learner’s return.

109. Instruction alignment

For instruction alignment, summative assessment begins by stating the educational claim and the evidence that could reasonably support it. Ask what capability is being sampled, what conditions shape performance and what decision will follow. Avoid collecting information simply because it is easy to count or administratively familiar.

In the fictional eduKateSG classroom, Adrian compares the intended learning with the actual demand of the task. Jo records important support conditions and inspects patterns rather than treating one response as a complete portrait. When evidence is ambiguous, the next move is another discriminating task or a narrower claim, not greater certainty in the wording.

The action should remain proportionate. If the evidence reveals a prerequisite gap, repair it. If the task samples too little, gather more evidence. If the curriculum introduces a later demand without an earlier dependency, repair the sequence. If performance is secure only under familiar conditions, add delay or transfer before broadening the conclusion.

Finally, inspect what happens after the decision. Did teaching change? Did the learner improve? Did later performance confirm the inference? Did the curriculum make subsequent learning easier to build? Summative assessment becomes trustworthy when its representations remain correctable by the learner’s return.

110. Resource alignment

For resource alignment, summative assessment begins by stating the educational claim and the evidence that could reasonably support it. Ask what capability is being sampled, what conditions shape performance and what decision will follow. Avoid collecting information simply because it is easy to count or administratively familiar.

In the fictional eduKateSG classroom, Adrian compares the intended learning with the actual demand of the task. Jo records important support conditions and inspects patterns rather than treating one response as a complete portrait. When evidence is ambiguous, the next move is another discriminating task or a narrower claim, not greater certainty in the wording.

The action should remain proportionate. If the evidence reveals a prerequisite gap, repair it. If the task samples too little, gather more evidence. If the curriculum introduces a later demand without an earlier dependency, repair the sequence. If performance is secure only under familiar conditions, add delay or transfer before broadening the conclusion.

Finally, inspect what happens after the decision. Did teaching change? Did the learner improve? Did later performance confirm the inference? Did the curriculum make subsequent learning easier to build? Summative assessment becomes trustworthy when its representations remain correctable by the learner’s return.

111. A six-week implementation

For a six-week implementation, summative assessment begins by stating the educational claim and the evidence that could reasonably support it. Ask what capability is being sampled, what conditions shape performance and what decision will follow. Avoid collecting information simply because it is easy to count or administratively familiar.

In the fictional eduKateSG classroom, Adrian compares the intended learning with the actual demand of the task. Jo records important support conditions and inspects patterns rather than treating one response as a complete portrait. When evidence is ambiguous, the next move is another discriminating task or a narrower claim, not greater certainty in the wording.

The action should remain proportionate. If the evidence reveals a prerequisite gap, repair it. If the task samples too little, gather more evidence. If the curriculum introduces a later demand without an earlier dependency, repair the sequence. If performance is secure only under familiar conditions, add delay or transfer before broadening the conclusion.

Finally, inspect what happens after the decision. Did teaching change? Did the learner improve? Did later performance confirm the inference? Did the curriculum make subsequent learning easier to build? Summative assessment becomes trustworthy when its representations remain correctable by the learner’s return.

112. Measuring quality

For measuring quality, summative assessment begins by stating the educational claim and the evidence that could reasonably support it. Ask what capability is being sampled, what conditions shape performance and what decision will follow. Avoid collecting information simply because it is easy to count or administratively familiar.

In the fictional eduKateSG classroom, Adrian compares the intended learning with the actual demand of the task. Jo records important support conditions and inspects patterns rather than treating one response as a complete portrait. When evidence is ambiguous, the next move is another discriminating task or a narrower claim, not greater certainty in the wording.

The action should remain proportionate. If the evidence reveals a prerequisite gap, repair it. If the task samples too little, gather more evidence. If the curriculum introduces a later demand without an earlier dependency, repair the sequence. If performance is secure only under familiar conditions, add delay or transfer before broadening the conclusion.

Finally, inspect what happens after the decision. Did teaching change? Did the learner improve? Did later performance confirm the inference? Did the curriculum make subsequent learning easier to build? Summative assessment becomes trustworthy when its representations remain correctable by the learner’s return.

113. Failure: score as learner

For failure: score as learner, summative assessment begins by stating the educational claim and the evidence that could reasonably support it. Ask what capability is being sampled, what conditions shape performance and what decision will follow. Avoid collecting information simply because it is easy to count or administratively familiar.

In the fictional eduKateSG classroom, Adrian compares the intended learning with the actual demand of the task. Jo records important support conditions and inspects patterns rather than treating one response as a complete portrait. When evidence is ambiguous, the next move is another discriminating task or a narrower claim, not greater certainty in the wording.

The action should remain proportionate. If the evidence reveals a prerequisite gap, repair it. If the task samples too little, gather more evidence. If the curriculum introduces a later demand without an earlier dependency, repair the sequence. If performance is secure only under familiar conditions, add delay or transfer before broadening the conclusion.

Finally, inspect what happens after the decision. Did teaching change? Did the learner improve? Did later performance confirm the inference? Did the curriculum make subsequent learning easier to build? Summative assessment becomes trustworthy when its representations remain correctable by the learner’s return.

114. Failure: task-goal mismatch

For failure: task-goal mismatch, summative assessment begins by stating the educational claim and the evidence that could reasonably support it. Ask what capability is being sampled, what conditions shape performance and what decision will follow. Avoid collecting information simply because it is easy to count or administratively familiar.

In the fictional eduKateSG classroom, Adrian compares the intended learning with the actual demand of the task. Jo records important support conditions and inspects patterns rather than treating one response as a complete portrait. When evidence is ambiguous, the next move is another discriminating task or a narrower claim, not greater certainty in the wording.

The action should remain proportionate. If the evidence reveals a prerequisite gap, repair it. If the task samples too little, gather more evidence. If the curriculum introduces a later demand without an earlier dependency, repair the sequence. If performance is secure only under familiar conditions, add delay or transfer before broadening the conclusion.

Finally, inspect what happens after the decision. Did teaching change? Did the learner improve? Did later performance confirm the inference? Did the curriculum make subsequent learning easier to build? Summative assessment becomes trustworthy when its representations remain correctable by the learner’s return.

115. Failure: feedback too late

For failure: feedback too late, summative assessment begins by stating the educational claim and the evidence that could reasonably support it. Ask what capability is being sampled, what conditions shape performance and what decision will follow. Avoid collecting information simply because it is easy to count or administratively familiar.

In the fictional eduKateSG classroom, Adrian compares the intended learning with the actual demand of the task. Jo records important support conditions and inspects patterns rather than treating one response as a complete portrait. When evidence is ambiguous, the next move is another discriminating task or a narrower claim, not greater certainty in the wording.

The action should remain proportionate. If the evidence reveals a prerequisite gap, repair it. If the task samples too little, gather more evidence. If the curriculum introduces a later demand without an earlier dependency, repair the sequence. If performance is secure only under familiar conditions, add delay or transfer before broadening the conclusion.

Finally, inspect what happens after the decision. Did teaching change? Did the learner improve? Did later performance confirm the inference? Did the curriculum make subsequent learning easier to build? Summative assessment becomes trustworthy when its representations remain correctable by the learner’s return.

116. Failure: over-testing

For failure: over-testing, summative assessment begins by stating the educational claim and the evidence that could reasonably support it. Ask what capability is being sampled, what conditions shape performance and what decision will follow. Avoid collecting information simply because it is easy to count or administratively familiar.

In the fictional eduKateSG classroom, Adrian compares the intended learning with the actual demand of the task. Jo records important support conditions and inspects patterns rather than treating one response as a complete portrait. When evidence is ambiguous, the next move is another discriminating task or a narrower claim, not greater certainty in the wording.

The action should remain proportionate. If the evidence reveals a prerequisite gap, repair it. If the task samples too little, gather more evidence. If the curriculum introduces a later demand without an earlier dependency, repair the sequence. If performance is secure only under familiar conditions, add delay or transfer before broadening the conclusion.

Finally, inspect what happens after the decision. Did teaching change? Did the learner improve? Did later performance confirm the inference? Did the curriculum make subsequent learning easier to build? Summative assessment becomes trustworthy when its representations remain correctable by the learner’s return.

117. Failure: under-sampling

For failure: under-sampling, summative assessment begins by stating the educational claim and the evidence that could reasonably support it. Ask what capability is being sampled, what conditions shape performance and what decision will follow. Avoid collecting information simply because it is easy to count or administratively familiar.

In the fictional eduKateSG classroom, Adrian compares the intended learning with the actual demand of the task. Jo records important support conditions and inspects patterns rather than treating one response as a complete portrait. When evidence is ambiguous, the next move is another discriminating task or a narrower claim, not greater certainty in the wording.

The action should remain proportionate. If the evidence reveals a prerequisite gap, repair it. If the task samples too little, gather more evidence. If the curriculum introduces a later demand without an earlier dependency, repair the sequence. If performance is secure only under familiar conditions, add delay or transfer before broadening the conclusion.

Finally, inspect what happens after the decision. Did teaching change? Did the learner improve? Did later performance confirm the inference? Did the curriculum make subsequent learning easier to build? Summative assessment becomes trustworthy when its representations remain correctable by the learner’s return.

118. Failure: curriculum coverage only

For failure: curriculum coverage only, summative assessment begins by stating the educational claim and the evidence that could reasonably support it. Ask what capability is being sampled, what conditions shape performance and what decision will follow. Avoid collecting information simply because it is easy to count or administratively familiar.

In the fictional eduKateSG classroom, Adrian compares the intended learning with the actual demand of the task. Jo records important support conditions and inspects patterns rather than treating one response as a complete portrait. When evidence is ambiguous, the next move is another discriminating task or a narrower claim, not greater certainty in the wording.

The action should remain proportionate. If the evidence reveals a prerequisite gap, repair it. If the task samples too little, gather more evidence. If the curriculum introduces a later demand without an earlier dependency, repair the sequence. If performance is secure only under familiar conditions, add delay or transfer before broadening the conclusion.

Finally, inspect what happens after the decision. Did teaching change? Did the learner improve? Did later performance confirm the inference? Did the curriculum make subsequent learning easier to build? Summative assessment becomes trustworthy when its representations remain correctable by the learner’s return.

119. Failure: disconnected units

For failure: disconnected units, summative assessment begins by stating the educational claim and the evidence that could reasonably support it. Ask what capability is being sampled, what conditions shape performance and what decision will follow. Avoid collecting information simply because it is easy to count or administratively familiar.

In the fictional eduKateSG classroom, Adrian compares the intended learning with the actual demand of the task. Jo records important support conditions and inspects patterns rather than treating one response as a complete portrait. When evidence is ambiguous, the next move is another discriminating task or a narrower claim, not greater certainty in the wording.

The action should remain proportionate. If the evidence reveals a prerequisite gap, repair it. If the task samples too little, gather more evidence. If the curriculum introduces a later demand without an earlier dependency, repair the sequence. If performance is secure only under familiar conditions, add delay or transfer before broadening the conclusion.

Finally, inspect what happens after the decision. Did teaching change? Did the learner improve? Did later performance confirm the inference? Did the curriculum make subsequent learning easier to build? Summative assessment becomes trustworthy when its representations remain correctable by the learner’s return.

120. Failure: hidden prerequisites

For failure: hidden prerequisites, summative assessment begins by stating the educational claim and the evidence that could reasonably support it. Ask what capability is being sampled, what conditions shape performance and what decision will follow. Avoid collecting information simply because it is easy to count or administratively familiar.

In the fictional eduKateSG classroom, Adrian compares the intended learning with the actual demand of the task. Jo records important support conditions and inspects patterns rather than treating one response as a complete portrait. When evidence is ambiguous, the next move is another discriminating task or a narrower claim, not greater certainty in the wording.

The action should remain proportionate. If the evidence reveals a prerequisite gap, repair it. If the task samples too little, gather more evidence. If the curriculum introduces a later demand without an earlier dependency, repair the sequence. If performance is secure only under familiar conditions, add delay or transfer before broadening the conclusion.

Finally, inspect what happens after the decision. Did teaching change? Did the learner improve? Did later performance confirm the inference? Did the curriculum make subsequent learning easier to build? Summative assessment becomes trustworthy when its representations remain correctable by the learner’s return.

121. Failure: false precision

For failure: false precision, summative assessment begins by stating the educational claim and the evidence that could reasonably support it. Ask what capability is being sampled, what conditions shape performance and what decision will follow. Avoid collecting information simply because it is easy to count or administratively familiar.

In the fictional eduKateSG classroom, Adrian compares the intended learning with the actual demand of the task. Jo records important support conditions and inspects patterns rather than treating one response as a complete portrait. When evidence is ambiguous, the next move is another discriminating task or a narrower claim, not greater certainty in the wording.

The action should remain proportionate. If the evidence reveals a prerequisite gap, repair it. If the task samples too little, gather more evidence. If the curriculum introduces a later demand without an earlier dependency, repair the sequence. If performance is secure only under familiar conditions, add delay or transfer before broadening the conclusion.

Finally, inspect what happens after the decision. Did teaching change? Did the learner improve? Did later performance confirm the inference? Did the curriculum make subsequent learning easier to build? Summative assessment becomes trustworthy when its representations remain correctable by the learner’s return.

122. Failure: no transfer

For failure: no transfer, summative assessment begins by stating the educational claim and the evidence that could reasonably support it. Ask what capability is being sampled, what conditions shape performance and what decision will follow. Avoid collecting information simply because it is easy to count or administratively familiar.

In the fictional eduKateSG classroom, Adrian compares the intended learning with the actual demand of the task. Jo records important support conditions and inspects patterns rather than treating one response as a complete portrait. When evidence is ambiguous, the next move is another discriminating task or a narrower claim, not greater certainty in the wording.

The action should remain proportionate. If the evidence reveals a prerequisite gap, repair it. If the task samples too little, gather more evidence. If the curriculum introduces a later demand without an earlier dependency, repair the sequence. If performance is secure only under familiar conditions, add delay or transfer before broadening the conclusion.

Finally, inspect what happens after the decision. Did teaching change? Did the learner improve? Did later performance confirm the inference? Did the curriculum make subsequent learning easier to build? Summative assessment becomes trustworthy when its representations remain correctable by the learner’s return.

123. Research base

For research base, summative assessment begins by stating the educational claim and the evidence that could reasonably support it. Ask what capability is being sampled, what conditions shape performance and what decision will follow. Avoid collecting information simply because it is easy to count or administratively familiar.

In the fictional eduKateSG classroom, Adrian compares the intended learning with the actual demand of the task. Jo records important support conditions and inspects patterns rather than treating one response as a complete portrait. When evidence is ambiguous, the next move is another discriminating task or a narrower claim, not greater certainty in the wording.

The action should remain proportionate. If the evidence reveals a prerequisite gap, repair it. If the task samples too little, gather more evidence. If the curriculum introduces a later demand without an earlier dependency, repair the sequence. If performance is secure only under familiar conditions, add delay or transfer before broadening the conclusion.

Finally, inspect what happens after the decision. Did teaching change? Did the learner improve? Did later performance confirm the inference? Did the curriculum make subsequent learning easier to build? Summative assessment becomes trustworthy when its representations remain correctable by the learner’s return.

124. The final mechanism

For the final mechanism, summative assessment begins by stating the educational claim and the evidence that could reasonably support it. Ask what capability is being sampled, what conditions shape performance and what decision will follow. Avoid collecting information simply because it is easy to count or administratively familiar.

In the fictional eduKateSG classroom, Adrian compares the intended learning with the actual demand of the task. Jo records important support conditions and inspects patterns rather than treating one response as a complete portrait. When evidence is ambiguous, the next move is another discriminating task or a narrower claim, not greater certainty in the wording.

The action should remain proportionate. If the evidence reveals a prerequisite gap, repair it. If the task samples too little, gather more evidence. If the curriculum introduces a later demand without an earlier dependency, repair the sequence. If performance is secure only under familiar conditions, add delay or transfer before broadening the conclusion.

Finally, inspect what happens after the decision. Did teaching change? Did the learner improve? Did later performance confirm the inference? Did the curriculum make subsequent learning easier to build? Summative assessment becomes trustworthy when its representations remain correctable by the learner’s return.

110. Original story: The Number and the Folder

Ben’s report showed 72. His folder showed something the number could not: strong algebra, fragile geometry selection, accurate untimed work and unfinished final questions under time pressure.

The score was real. So were its boundaries. Adrian used the result for the decision it was designed to support and the item pattern for the next learning plan.

111. Summative judgment must remain bounded

A summative result compresses sampled performance into a usable judgment. Compression is necessary for many institutional decisions, but responsible interpretation preserves what the score does not contain.

112. Connected routes

Use Learning Goals for alignment, Formative Assessment for in-process adaptation and Diagnostic Teaching when the result must be unpacked for repair.

113. A score is a claim with conditions

Good summative assessment makes a useful judgment without pretending the judgment is the whole person, the whole curriculum or the whole future. Its authority grows when its limits remain visible.