EDUKATE SINGAPORE · GOLD STANDARD REFERENCE PROGRAM · VOLUME 28
Keep decisions intelligible and temporary departures bounded, traceable and reviewable.
Choose a reading route, consult the chapter index, or begin with Chapter 1.
Full chapter index · Practical exercise · Reference-programme reading sequence
A decision record explains what was decided, why, by whom and under which conditions. An exception register keeps departures from the usual method visible until they are resolved, expired or replaced by an authorised change. In the Gold Standard programme, these records help a team act consistently without pretending that every situation is identical.
This volume provides a practical method for decisions that need to remain intelligible after the meeting ends. It does not require a lengthy report for every routine choice. Its focus is the decision whose reasoning, authority or continuing conditions matter to another person. The running example is a fictional community workshop considering temporary use of a substitute activity kit.
Read Governance, Ownership and Decision Rights for the wider authority model and Information and Record Integrity for record quality. This volume applies those ideas to decision history and temporary departures. It adds no new governing requirement and does not authorise exceptions to external obligations.
Chapter index
Chapters 1–2: Frame the decision
Chapters 3–4: Build the reasoning
Chapters 5–6: Authorise the departure
Chapters 7–8: Bound and track
Chapters 9–10: Learn and practise
A durable record is useful when another person must act on a decision, when conditions could change its validity or when future review needs to understand the reasoning. It is especially useful for unusual acceptance, changed priorities, temporary methods and decisions made with important uncertainty.
Routine choices can use a defined operating rule and an ordinary transaction record. The aim is not to slow every task with a separate approval essay. The record should be proportionate to the consequence, novelty and difficulty of reconstructing the decision later.
At the workshop, selecting a kit already approved for the session may be routine. Using an unfamiliar substitute because the usual kit is unavailable needs closer examination. The substitute may affect the promised activity, preparation method or competence needed to run it.
Recording test: If the decision-maker were unavailable next week, could the next responsible person identify the permitted action and its limits? If the answer depends on remembering a conversation, create a record at the level of detail the decision actually requires.
CHAPTER 2 OF 10 · Frame the decision
2. State the question before recording the answer
A decision record should begin with a question that someone has authority to answer. “Kit issue” is not a decision question. “May this substitute be used for Friday's stated activity after the listed checks?” identifies an action, scope and condition.
Separate the desired outcome from a preferred solution. The outcome might be a usable session. The proposed substitute is one way to attempt it. Other options could include a different activity, an agreed reschedule or declining to proceed under the current conditions.
Identify the decision deadline and why it matters. A deadline can affect which evidence is obtainable, but it does not make missing evidence disappear. Where a safe, authorised decision cannot be reached in time, the consequence may be a narrower commitment or a stop.
Question check: Could two readers tell whether the final answer actually addresses the original question? If the record shifts from approving one session to approving all future substitutes, the scope has changed and needs an explicit decision.
CHAPTER 3 OF 10 · Build the reasoning
3. Preserve evidence and uncertainty separately
Record the observations that matter and where they can be examined. Then explain the interpretation. A photograph may show the kit's identity or visible condition; it does not necessarily show that the kit supports the intended activity. A successful demonstration supports the conditions actually demonstrated.
Keep missing information visible. Use a statement such as “No evidence yet for use with this group size” rather than disguising the gap as an optimistic assumption. An assumption may still inform planning, but it should have a label and a consequence if wrong.
The workshop can record that a substitute was checked against the intended activity and that the operator demonstrated the revised method. It should also record any part that was not examined. That boundary protects future readers from enlarging the decision beyond its evidence.
Evidence chain: Connect requirement, observation, interpretation, decision and follow-up. If a conclusion cannot be traced through that chain, identify the missing link. A reference to a large folder is not enough when nobody can tell which item supports the important claim.
List the alternatives that could genuinely affect the choice. Do not fill the record with obviously impossible options merely to make the preferred answer look considered. State the relevant benefit, cost, limitation and effect on existing commitments for each plausible route.
A comparison need not always use scores. A critical missing condition can rule out an option even when it performs well on several less important features. Do not average away a requirement that must be met.
| Option | What it could preserve | What must be checked |
|---|---|---|
| Use the proposed substitute | The planned session time | Suitability, readiness and authorised scope |
| Change the activity | A useful session with a different outcome | The participant's understanding and the revised method |
| Reschedule by agreement | The original activity | A genuine available date and effects on commitments |
| Do not proceed | Avoid an unsupported promise | Clear communication and treatment of open responsibilities |
Record why an alternative was rejected when that reason may matter later. A route unavailable today may become viable after a resource or condition changes. The history should help a later decision-maker reconsider it without reconstructing the entire discussion.
A decision should identify who authorised it and the basis of that authority. Attendance at a meeting does not automatically make every attendee a decision-maker. Nor does a senior title necessarily confer authority over every kind of departure.
Write the permitted action in operational language. Include the affected scope, required checks, responsible operator and any event that stops or invalidates the permission. If the decision is “proceed only after verification,” distinguish that conditional permission from evidence that verification has happened.
Make communication part of execution. The person who must act needs the decision and its limits; affected people need a truthful explanation of any changed promise. A perfectly written record that remains in the approver's folder does not control the work.
Release check: Before action begins, can the operator state what is allowed, what remains prohibited, which conditions have been met and where to escalate a change? If those answers differ from the record, repair the handoff.
CHAPTER 6 OF 10 · Authorise the departure
6. Distinguish an exception from a change
An exception is a bounded departure from the usual method for a stated case or period. A controlled change alters the method that will govern future work. Confusing them allows a temporary arrangement to become a permanent standard without examination.
In the fictional workshop, using one checked substitute for one session might be an exception to a local preference for the usual kit. Adopting that substitute across all sessions is a different decision with a wider evidence burden.
Some boundaries cannot be waived by the local decision-maker. This manual does not grant authority to bypass external obligations, essential acceptance conditions or the programme's conformance gate. Where a departure means an applicable requirement is not met, preserve that finding. Calling the departure approved does not make the requirement met.
Classification test: Ask whether the normal method remains appropriate. If the same exception is repeatedly needed, examine whether the method, resources or promise should change. Repetition is a review signal, not automatic approval for continued departure.
CHAPTER 7 OF 10 · Bound and track
7. Keep temporary permission from becoming indefinite
An exception needs a clear scope, owner, review trigger and ending condition. “Until further notice” may leave a temporary arrangement active long after its assumptions have changed. Use a meaningful time or event boundary and specify what happens when it is reached.
The end of an exception does not automatically mean that returning to the old method is ready or safe. Review the actual work and choose an authorised next state: resume the established method, approve a controlled change, narrow the scope or stop the affected activity.
If an exception must be reconsidered, review current evidence rather than copying the previous approval. Identify whether the exposure has accumulated, whether support remains available and whether the promised corrective action happened.
Expiry question: Who will notice that the permission is ending, and what action will they take? A date in a register is not a control unless a responsible person uses it to make the next decision.
A register should make current decisions findable without turning every row into a full report. Store the essential status and link to the detailed evidence where needed. Use stable identities so related cases and later changes can be connected.
A practical entry includes the decision question, governing method, departure, reason, evidence, owner, approver, scope, start, review trigger, ending condition and current status. Select the fields necessary for the actual work; this is a programme template, not a universal external form.
Keep status meanings clear. Proposed, approved with conditions, active, suspended, expired and closed describe different events. An approved exception may not yet be active. A closed decision may still have improvement work elsewhere, which needs a linked owner.
Register test: Ask someone to identify all active departures affecting the next session. If they must read every historical file or guess which “approved” rows still apply, improve the current-state view while preserving the history.
CHAPTER 9 OF 10 · Learn and practise
9. Review outcomes and preserve decision history
After the relevant work, compare what happened with the decision's assumptions and conditions. Was the permitted action performed? Were required checks completed? Did an unexpected condition appear? Did the outcome support, narrow or contradict the original reasoning?
A good result does not automatically prove the decision process was sound. The team may have been fortunate or may have relied on support not recorded in the plan. Conversely, an adverse result does not by itself show that every earlier judgement was unreasonable. Examine the evidence available and the control of the actual execution.
Correct records transparently. Preserve the earlier decision and link an amended or superseding decision with its reason. Do not rewrite the original as though the later evidence had always been known. That destroys the history needed for learning.
For changes that will affect future methods, use Change, Scaling and Renewal. A completed exception can reveal a better method, but that method still needs a controlled adoption decision.
Case: A substitute kit worked in one small workshop with the outgoing coordinator present. The team now wants to use it in every session for the next month. The register says “approved,” with no scope, expiry or record of the coordinator's assistance. A new operator assumes the substitute is the normal method.
Reasoned response: Recover the actual scope and evidence of the earlier decision without inventing missing facts. Record uncertainty where the history cannot be reconstructed. The proposed month-long use and new operator are changed conditions. Examine readiness for that scope, establish authority and communicate the current permitted method before relying on the earlier label.
Write a compact decision record containing the question, observations, uncertainties, alternatives, authorised action, limits and review trigger. Then create a register row that allows another operator to find the current decision. Include the action required when the permission ends.
The exercise applies existing requirements for criteria, authority, controlled execution, evidence, exceptions and change. It does not waive those requirements. The practical goal is simple: a future reader should understand both what the organisation chose and why that choice was considered supportable at the time.
Contents · Previous chapter · Continue through the reference program
The Gold Standard reference program · 32 volumes
Begin with the four foundation volumes, then choose the evidence, operating, assessment or study guides that match your task. These reading routes extend the manual without changing its twelve core requirements.
Volumes 1–4 · Foundations
Volumes 5–8 · Evidence and governance
Volumes 9–12 · People and operating systems
Volumes 13–16 · Assurance and assessment
Volumes 17–20 · Workbooks and reference tools
Volumes 21–24 · Advanced applications
Volumes 25–28 · Everyday management controls
Volumes 29–32 · Study and teaching guides
Return to The Gold Standard master hub
Apply the framework to a specialist field
- Education Management — connect the parts of a learning system.
- Learning Assessment — examine what learning evidence can establish.
- Operations Management — follow plans into everyday delivery.
- Business Continuity — examine continuity under disruption.
Edition, sources and interpretation
The definitions, GS-R01 to GS-R12, assessment findings, teaching loops, templates and worked cases are the original editorial synthesis of the eduKate Gold Standard Reference Program. Examples do not report audited organisational results. This volume was prepared on 14 October 2026 and applies programme Edition 1.0, established on 13 October 2026. A future change to requirements should identify the affected edition and preserve the interpretation of earlier evidence.
The working definitions and requirements in this volume are program-specific. For an externally governed task, consult its applicable authoritative vocabulary and technical requirements directly. No numbered clause here is presented as a quotation or reproduction of an external standard.
