Jump to text
On this page
Copy and source
Published source confirmed at last check

Source changed 2026-11-18 21:11:39 UTC · snapshot created 2026-11-18 21:12:35 UTC · last check 2026-11-18 22:15:20 UTC

NSTD.6 - Declared-Use Narrative Rendering Quality Evaluation

Status: Stable

Type: DPF evaluation pattern body

Primary EntityOfConcern: NarrativeRenderingQualityEvaluationCharacteristicSpace@Context, the characteristics, scales and evidence used to evaluate a narrative rendering for a specified use.

NSTD.6:1 - Problem frame

Use this pattern to assess a version of a narrative for a particular reader or listener and use. An editor can inspect an existing handout or an unaccepted draft. Decide whether the question concerns the present product, its fidelity to a specified source, or what the author did to produce or repair it.

The evaluated object, NarrativeRenderingVersion@Context, is an identifiable version of a narrative account as expressed for a reader. A script, a slide sequence or a generated draft may contain one. Identify the account within a collection or teaching programme. An unknown source or a defect can itself be an evaluation finding. Assess the authority of the collection, the reliability of its generator or the quality of the entire programme under the criteria for that object.

Identify the version, its reader and the work the account should support. Select the relations and characteristics needed for that work. Check that the criteria distinguish an adequate account from a deficient one and refer a question about another object to the appropriate evaluation.

A technical narrative can be easy to read and still turn precedence into cause or omit the condition under which a reader must consult the source.

A repeatable evaluation identifies the grounds for each judgement. Use it in E.23 when the task includes improving the narrative.

Select the characteristics below for the use in question. They cover source selection, structural recovery, order, time and roles, source return and engagement. Examine evidence, assurance, ethics, publication authority or historical work when the corresponding claim is at stake.

NSTD.6:2 - Problem

A fluent narrative can misstate a relation. An accurate one can leave learners unable to follow it. To judge a version for its use, specify the object, characteristics, scales, evidence, treatment of missing information and thresholds. Show what supports each value, what an exceptional result would establish and what could repair a consequential deficiency.

NSTD.6:3 - Forces

ForceTension
Fluency and structural recoveryA reader can follow the prose while missing a relation needed for the task.
Engagement and informed useA compelling story can conceal an uncertainty that changes the reader’s choice.
Local use and a reusable scaleA short rubric may serve one project. To compare judgements across cases, evaluators need common value meanings and stated conditions.
Measurement vs evaluationSome values may be measured through C.16; many are ordinal content evaluations.
Improvement and proxy measuresImproving a measured indicator can damage a quality the indicator omits. Track the consequential effects on those qualities as well.

NSTD.6:4 - Solution

Use an existing evaluation result if it answers the question for this narrative version, reader and use. If a suitable evaluation specification is available, apply it to the version. Develop or adapt a specification for any remaining question about which characteristics to judge and what evidence counts.

For explanatory technical prose, apply F.19 to the connected text. Recover its governing message, compare it with the selected source and test the contribution of each optional contrast, caveat or extra proposition for this reader and genre. Use F.19:4.4 for a proposed warning and F.19:4.5 for the reading evidence the conclusion needs. This wording judgement contributes to LanguageStatePrecisionAndCoarseningFit.

Decide what counts as a supported answer. Compare a factual account with evidence and a literary reading with the text and serious alternatives. For a deliberately open story, the reader may need to explain why several interpretations fit. For a fictional construction or a hypothetical real scenario, inspect the premises and their consequences. If the question concerns audience experience, examine evidence of that response. Use NSTD.9 for an interpretation still needed and NSTD.10 to construct missing events. Comparing two expressions can establish fidelity. Whether the source was adequate, the new account was worth producing or the audience responded as intended requires grounds for that claim.

A faithful story can show every step of a particular run. If the author was asked to explain how to design a procedure for a range of situations, use EXD.Reference:3 and EXD.6:4.1 to see whether the finished story lets the reader produce and justify a suitable design. Several designs may be admissible. Give the story credit for the decisions and grounds its reader can recover.

Use NarrativeRenderingEpiplexity to assess how much selected structure a reader can correctly extract. It specializes FPF C.2.8 ExtractableStructuralInformation. Identify the receiving narrative episteme (the account conveyed), the particular expression conveying it and the reader or observer. Specify the source structure, correctness criterion, usable prior knowledge, operations, help, access and budget. A rendering reference can identify both account and expression when it is unambiguous. Record the conditions under the basis, scale and evidence fields below.

If naming the additional and missing relations answers the question, report them. For a count, define the units, then count the correctly recovered selected semantic relations. A fraction n/N needs a fixed nonempty selected set. For each unit preserve the participants, predicate, polarity, modality and conditions that change use. Count a relation once however often it is repeated or depicted. Keep its identity with the count. Two readers may each recover four of five relations while only one recovers the condition for stopping. Give any ordinal or nonnegative weight its structural meaning. Score utility separately. Use the formal mapping in C.2.8 for a bit estimate.

State whether the basis is a conditional walkthrough, an observed reading or a formal estimate with a specified mapping. A first response under known conditions shows the recovery observed in that trial and can inform an estimate of extractability. Establish a maximum, impossibility or reliability across a class only with evidence covering that claim. For a computational amount, specify the method and model. Retain the first response, help used and source returns. A familiar relation can still be recoverable from the expression. If an expert supplies a correction from prior knowledge, attribute that correction to its source. Decide whether correctness means recovering the source’s claim or recovering a warranted relation in the subject.

To isolate the effect of expression, hold the expressed claims and the other relevant conditions constant. A substantive change can identify another episteme under C.2.1. Report any changed access, preparation or help with the comparison it qualifies. Report performance after a perturbation or transfer under the new conditions. Use C.11.CRC and C.11 to compare effort and usefulness when relevant. A sufficient existing rendering can stay in use.

NarrativeRenderingQualityEvaluationCharacteristicSpace@Context:
  evaluatedObjectKindRef: NarrativeRenderingVersion@Context
  declaredUseScope:
  objectKindFitRule:
  discriminatingCaseSet:
  characteristicSlotSet:
  epiplexityBasisRule:
  scaleBindingSet:
  valueMeaningSet:
  evidenceBasisRule:
  missingnessAndLoweringRule:
  resultRowShape:
  floorAndExceptionalMeaning:
  protectedTradeoffSet:
  stopOrReopenCondition:
  neighboringGoverningPatternRefs:

Identify the object before assigning a value.

Object-kind caseHandling
Narrative account, whether candidate or acceptedEvaluate every characteristic whose failure would defeat the intended use.
Narrative rendering below a required floorReport the supported value, the loss for this use and the smallest repair that could address it, if one is available.
The question concerns another objectEvaluate that object under its own criteria, such as reliability across a generator’s operating range or authenticity of a source archive.
The object proves to be another kind after the evaluation has begunState the mismatch and stop assigning values to this object under the narrative criteria.

Use these default meanings for the ordinal content-quality characteristics. Assign NarrativeRenderingEpiplexity its own structural amount and use condition as specified above.

ValueMeaning
0No usable basis for this quality judgement, or evaluation stopped before a quality value could be used. Give the cause.
1Object is a narrative rendering but unusable for the declared use.
2Suitable for orientation. Consult the source before relying on this account.
3Locally usable with named limitations and repair obligations.
4Good for declared use with bounded losses and source return.
5Strong for the declared use. A reader can retrace the current source relation, the result under a meaningful change and the consequential boundary cases. When the use includes a historical claim, the evidence also makes the relevant construction or repair traceable.

For teaching or other reliance-bearing use, the default is that every selected relation necessary for that use must be recoverable with the stated support and evidence. Specify those relations at the grain the task needs. A required distinction and its stop condition may suffice. Judge the amount recovered and the use condition separately. Unrelated relations cannot make up for a missing necessary one.

The other characteristics essential to the use have a default floor of 4 on their own scales, including OrderingRecoverability and SourceReturnReadiness when essential. When the account demonstrates a route through a constraint-governed unfolding structure, the default floor of 4 applies to recovering the route and the source constraints the declared use needs. A local, low-risk orientation use may set these quality floors to 3 if it explicitly excludes the downstream uses the result cannot support. Select the structure before evaluating a narrower use. Removing a difficult relation after seeing the answer defines a different comparison.

Result-row shape:

NarrativeRenderingQualityResultRow@Context:
  narrativeRenderingVersionRef:
  declaredUseScopeRef:
  characteristicId:
  characteristicName:
  scaleRef:
  value:
  evidenceBasisRefs:
  missingnessClass:
  loweringReason:
  repairAction:
  professionalOrAuthorityRefs?:
  reopenCondition:

For repeated improvement, collect the relevant judgements in a result for the version and use being compared.

NarrativeRenderingQualityEvaluationResult@Context:
  evaluatedNarrativeRenderingVersionRef:
  declaredUseScopeRef:
  evaluationCharacteristicSpaceRef: NarrativeRenderingQualityEvaluationCharacteristicSpace@Context
  evaluationPurpose:
  evidenceBasisRefs:
  resultRows:
  protectedTradeoffSet:
  belowFloorRows:
  candidateImprovementProposalRows?:
  nonUseBoundary?:
  stopOrReopenCondition:

When repeated improvement is wanted, use E.23 with NSTD.6 as the evaluation of the narrative. Use E.22 first if the quality question still needs framing.

NarrativeRenderingImprovementLoopInput@Context:
  e22QuestionFrameRef?:
  objectUnderImprovementRef: NarrativeRenderingVersion@Context
  objectVersionBeforeRef:
  objectUnderImprovementEvaluationRef: NSTD.6
  improvementAim:
  protectedTradeoffSet:
  costAndRiskAccount:
  allowedChangeSlice:
    narrativeRenderingVersion | NSTD.1-intake | NSTD.2-ordering |
    NSTD.3-event-model | NSTD.4-viewpoint | NSTD.5-engagement |
    NSTD.7-generation-and-inspection | NSTD.8-learning-route |
    NSTD.9-interpretation | NSTD.10-story-construction |
    evaluationCharacteristicSpace
  returnedFindingOrProposalRows:
  expectedReEvaluationResultForm: NarrativeRenderingQualityEvaluationResult@Context
  neighboringGoverningPatternRefs:
  stopContinueSwitchOrHoldCondition:

To judge improvement, evaluate the changed version through NSTD.6 or a declared stronger evaluation. Compare it with an applicable baseline. Claim the gain that comparison establishes under the protected conditions. Specify what the change should improve and which qualities it must protect, including for a style edit, another generation attempt or added drama. If the change affects the source set or its currency, admission of generated material, the teaching or publication medium, an ethics, evidence or assurance claim, or the evaluation criteria, identify the applicable method. Include that work in the change being assessed or commission it separately. Use B.4 for a claim about evolution under repeated use. Under G.11, reconsider affected results when the source, reader observations, teaching evidence, generation practice or FPF edition changes enough to alter reliance. Reuse sufficient unaffected results.

Before assigning values, recover the source and use basis for the question being answered.

For current-product quality, derive the necessary structure from the work readers need to do, their prerequisites and qualified subject sources. Use the account to refine that selection and inspect what it conveys, hides or loses. Point to the relevant source and rendering passages. Recover the ordering rationale, conditions for source return and any consequential event, mechanism, viewpoint, agency, engagement or learning-route relations, using NSTD.1 through NSTD.5 and NSTD.8 as needed. Report this reconstruction as a present assessment with its uncertainties. Missing subject grounds, inaccessible returns and unsupported relations affect that assessment under their own criteria.

For a promised transformation, compare the named source and result against the preservation or change promised. For a claim about how the work was performed, use records or other evidence of the relevant actions. Trace the actual selection, order, mechanism, viewpoint, engagement and learning-route choices that the claim needs under NSTD.1 through NSTD.5 and NSTD.8. State which actions the evidence establishes. If the history remains unknown, give that limit to the work relying on it.

Evaluate a candidate before acceptance when its quality is in question. The result may support acceptance, repair or rejection. Use C.35 for generated or discovered content intended to inform architecture work. Before reusing a narrative as evidence, establish the relevant subject grounds. Use A.10 when the source-to-use account is implicit or must remain reproducible. For assurance, authority to publish or an effect on human learning, obtain evidence or authorization appropriate to that claim.

Carry out the parts of the evaluation that answer the question.

  1. Identify the version and the particular account in its text, slides or other expression. Direct a question about the whole course, source collection or generator to the evaluation for that object.
  2. Select evidence for the claim. Inspect the subject and rendering for present quality, compare the named source and result for promised fidelity, or establish the claimed work from evidence of its performance. A single admissible route can show that an outcome is possible. One admissible route that misses a deadline refutes a claim that every admissible route meets it. A reader who must choose a route as conditions change needs the rule for that choice and a way to recognize when it applies. Assess that content even when the illustrated route is correct.
  3. Identify the work this reader or listener needs the account for and the conditions under which the judgement holds. State limits that affect reliance on the result.
  4. Evaluate every characteristic whose failure would defeat the declared use or make it unsafe. Select additional characteristics only when their result matters to the question.
  5. When a value falls below its floor, identify the loss. Give the smallest repair that would address it, or specify what is still needed to find one.
  6. Before claiming an improvement, re-evaluate the affected characteristics if the version or selected source basis changes.

Missingness and lowering rules:

Missing or defect conditionLowering rule
Selected structures cannot be identified from the task and subject sourcesLeave the structural amount unassigned. Use NSTD.1 to complete source selection, NSTD.9 if the needed interpretation is missing or disputed, or NSTD.10 if the proposed events still need construction. State any separate mismatch in the evaluated object.
The reader, correctness criterion or another consequential condition of comparison is unknownLeave the amount unassigned for that scope. Return any narrower finding that the available evidence supports and examine independently answerable questions.
Ordering rule absentOrderingRecoverability no higher than 2.
A missing distinction of time, viewpoint, participant or mediation changes the reading for this useSet TemporalPostureAndRoleFit no higher than 2. Use NSTD.1 or NSTD.4 to resolve that ambiguity. Apply the limit when the distinction matters to the interpretation.
A source-selection rationale or reader-use hypothesis needed for the evaluation cannot be recoveredSet SourceStructureSelectionFit no higher than 2 for that use and return the selection question to NSTD.1. Lower TemporalPostureAndRoleFit if its time or participant distinctions are also missing. Leave structural amount unassigned if its own selection, correctness or observer basis is unavailable. Check what can be recovered from all the pertinent material.
Source-return condition absentSourceReturnReadiness no higher than 2.
The account presents one route as obligatory although the governing conditions allow other routes, or omits a branch, return or exit that would change the reader’s conclusion or actionSet DemonstrativeSliceRecoverability no higher than 2 for that use. Restore the missing choice and governing condition through NSTD.1, NSTD.2 or A.22.CGUS. Accept an admissible route as a sufficient demonstration when the claim only requires one such route.
Simplification, literary wording or drama obscures a source relation, distinction of precision or limit of a coarser account needed for this useSet LanguageStatePrecisionAndCoarseningFit no higher than 2. Repair the specific loss when its meaning is known. Use C.2.LS, A.6.3.CSC, E.17.EFP, E.10, A.6.P or C.16.Q for a remaining question of language state, coarsening, explanation or precision, then reassess the affected characteristic.
Only an intended effect or seed is available, without an identifiable narrative accountNo narrative version is yet available for this evaluation. Develop the cue through A.16.1, NSTD.1 or NSTD.2, or evaluate the cue for its own use.
Persuasion conceals an uncertainty, risk or alternative material to the reader’s choiceSet EngagementBoundedness no higher than 3. Restore the relevant condition or alternative through NSTD.5. Use the applicable D.1–D.5 contribution if a consequential value or harm question remains for an affected party.
A generated narrative is still a candidateEvaluate it for the stated use. Record the basis and limits. Use NSTD.7 when the generation method needs repair and C.35 when the content is meant to inform architecture work.
A claim about evidence or assurance lacks groundsState which claim is unsupported and recover its criterion. Use A.10 for a source-to-use account that is implicit or must be reproducible and B.3 for a specified assurance claim. Judge each quality under its own criterion. Leave an unsupported structural amount unassigned.
The present product can be assessed, but its construction or repair history is unknownUse the available product evidence for current quality. Report the historical limit to whoever relies on that claim and apply the relevant source, use and admission criteria to their own questions.

A defined trial in which the reader recovers none of the selected relations supports observed recovery 0 for that trial. An empty selected set has no recovery fraction. Identify the object before assigning an amount. A walkthrough can support a conditional estimate. Use an existing reading when it answers an empirical question, or consider a new trial when the obtainable evidence would justify its cost for the task.

Default narrative rendering quality characteristics:

CharacteristicEvaluation questionLow-value repair action
SourceStructureSelectionFitWhat does the reader need from the source? What grounds support the selection and any assumption about reader interest?Use NSTD.1 to reconstruct or revise a deficient selection before changing the style, drama or generation instruction.
NarrativeRenderingEpiplexityWhich selected relations can this observer correctly extract from the account through this expression under the specified conditions? When comparing alternatives, which relations differ?Use C.2.8 for structural amount. Return a selection gap to NSTD.1. Repair a missing or false relation, its expression or a needed source return through the method for that problem. Reassess recovery if the repair changes what can be extracted. Use C.33 for an additional question about architecture-description adequacy.
OrderingRecoverabilityCan the reader recover the rationale for this order and the relevant relations the sequence leaves out?Use NSTD.2 to state the ordering rule and which relations it preserves or hides.
DemonstrativeSliceRecoverabilityCan the reader identify which route the account shows, why the source permits it, and which other possibilities matter for the intended inference or decision?Use NSTD.1, NSTD.2 or the source method to recover a constraint, alternative, loop or exit whose omission changes that use.
TemporalPostureAndRoleFitCan the reader distinguish the relevant time, mediation, participants and viewpoints?Use NSTD.1 or NSTD.4 to clarify observation, live reporting, proposal or fiction and who is speaking or acting. Obtain work-role, performer/assignment or architecture-admission evidence when the corresponding claim needs it.
EventMechanismSupportCan the reader reconstruct events, mechanisms, dependencies, or state changes when required?Reopen NSTD.3; add mechanism support or lower causal language.
ViewpointAgencyDisciplineDoes the viewpoint convey the source relation and accurately attribute action, capability, responsibility and permission?Use NSTD.4 to distinguish the protagonist or actant from whoever acted, held a role or a permission, or bore the relevant responsibility. Examine the ethical question when it arises.
EngagementBoundednessDoes engagement serve the intended effect while preserving source relations and the conditions needed for the reader’s choice?Reopen NSTD.5; restore the missing condition, revise the device or keep a sufficient simpler account.
LanguageStatePrecisionAndCoarseningFitDoes the chosen language state make the needed relations, precision and coarsening limits recoverable for this use? In explanatory technical prose, what does each optional rhetorical branch contribute to the intended reader?Apply the connected F.19 reading to explanatory technical prose, retaining the distinctions that change its use. For a remaining question, use C.2.LS for a threshold-bearing language-state profile, A.6.3.CSC for narrowed-use coarsening, E.17.EFP for explanation-facing retelling, A.16.1/A.16.2 for cue or backoff, and E.10, A.6.P or C.16.Q for precision restoration.
EthicsEvidenceAssuranceRoutingFor a consequential claim about value, harm, evidence, assurance or policy, are the grounds and limits available for this use?Identify the unresolved issue and use the pertinent D.1–D.5, A.10, B.3 or professional contribution.
MediumAndPublicationFitDoes the carrier fit the reader and use without changing the claim?Route publication or audience-unit questions to E.17, E.17.AUD, or NSTD.8.
SourceReturnReadinessDoes the narrative tell readers when and where to return to the source basis or relevant method?Add the missing return condition or destination, or narrow the supported use.
LearningRouteReconstructionFitWhat can the learner reconstruct under the intended support, and when must they return to a source?Use NSTD.8 to address an inadequate source contribution, support or next task. In a design, assess what the exercise affords. In a claim about a learner, use their response.

Use the ordinal meanings above and the adjacent-value examples below for LearningRouteReconstructionFit in teaching applications.

NSTD.6:5 - Archetypal Grounding

NSTD.6:5.1 - Value bank: full result rows

Each row evaluates an identifiable account for a particular use. Earlier acceptance is unnecessary, and a change in use can change the judgement.

CaseCharacteristicValueEvidence basisLow-value repair
FPF seminar handoutNarrativeRenderingEpiplexityQualitative illustration. The numerical amount is unknown.The case assumes that learners can recover the EntityOfConcern, forces, solution and relevant exits. Their names alone specify neither equal units nor a denominator or an observed trial.Retain this limited finding. If transfer matters, use evidence of a learner choosing an appropriate pattern in a different situation and report it under those conditions.
FPF seminar handoutSourceReturnReadiness4Each slogan in the constructed handout links to the relevant pattern. Before applying one in a new situation the exercise directs the reader to that pattern’s conditions. This shows when and where to return. A 5 also needs evidence under a meaningful change and at a consequential boundary.Reconsider this value if a change in task, preparation or access affects the return, or if a reader cannot find or use it. Repair a confirmed failure even if the source text is unchanged.
FPF seminar handoutEngagementBoundedness4This clearly fictional failure story invites readers to compare workloads and gives them the relevant conditions. The value is a judgement about the design.If increased attention is claimed, use an appropriate observed comparison. Restore any condition the persuasive use conceals.
Incomplete version of a homotopy lessonLanguageStatePrecisionAndCoarseningFit3The vivid analogy leaves readers to infer which continuous deformations are allowed and when to consult the formal account.Name the restriction on crossing the missing point and ask when it matters to contraction.
Homotopy explanationOrderingRecoverability4The explanation names its teaching order and tells readers where to recover the dependencies of the proof.To substantiate 5, examine a changed problem in which the reader maps the story order to the proof dependencies.
Franchise continuation probeSourceReturnReadiness4An illustrative repair of the Readme’s key case puts this return in the draft: “If a scene assumes access before Wednesday, consult the supplied premise that the character first obtains the key on Wednesday.” The condition and destination are explicit.If the premise or proposed access changes, reopen that continuity comparison. Check an appropriate changed case when robustness across cases matters. Reconsider new use conditions if they affect the return.
Live commentaryEventMechanismSupport3The commentator distinguishes observation from provisional interpretation. The listener receives only a general suggestion to consult later telemetry.Identify the official record, replay or statistic and when it could resolve the particular claim.
Generated graph-to-text narrativeEventMechanismSupport2The draft states “A caused B”, while the source graph establishes only that A preceded B.Return to NSTD.3 or NSTD.7: remove the unsupported cause or obtain the needed causal account.

NSTD.6:5.2 - An existing product and an unknown manufacturing history

An editor receives an onboarding handout with a qualified subject basis and source links. The record of its creation is missing. The editor identifies the relations needed for onboarding from the task and the subject sources, follows them through the present explanation and exercises, and tests a meaningful changed case. Inspection can establish a missing return or false relation. A strong boundary test can support a high present-product value. What the author did remains a historical question.

To assess fidelity to a particular seminar, the editor compares that seminar with the handout. To establish that the author reviewed or corrected it, the editor needs evidence of the action. A successful reader response bears on the reader’s performance. A generated draft can likewise receive a current-product evaluation before an acceptance decision or any historical investigation.

NSTD.6:5.3 - Before and after evaluation repair

Before evaluation statement:

The narrative is strong because readers liked it and remembered the main point.

The judgement asserts liking and recall of the main point. It leaves open the evidence for those claims, which source relations the readers recovered, what the relations support and which uses the result permits.

After evaluation statement:

For onboarding, assume learners can reconstruct the ordinary route for using the pattern. A numerical amount still needs a defined set of relations, conditions and evidence. Source-return readiness is 3 in this version. Readers can use the handout for the familiar case and know when to return to the source, but two slogans omit the references needed to locate the applicable conditions in a new case. Engagement boundedness is 4 because the constructed story invites comparison of workloads and preserves the conditions needed for that choice. If the task includes repair, add usable directions to those two sources first.

For an improvement under E.23, add those two references and compare whether readers can now locate the applicable conditions under the intended support. Revisit structural recovery if changed access or wording can affect it. Reuse a previous amount only with its selection, scale, conditions and evidence. A quality label of 4 provides no numerator or denominator.

NSTD.6:5.4 - Adjacent-value calibration

Fix five relations in an account of reusing a review after a change in display. The claims, the question and the qualification window must each be the same. If a required argument is unavailable, stop the new check and return to the source. This stop preserves unrelated review results. The example adapts C.37.CF. Count the condition, stop and return as one compound unit at this grain. Hold the reader, case facts, material, operations, help and budget constant. The three rows illustrate progressively larger constructed sets within this selection. For another selection, keep its own denominator and the identities of the relations.

Correctly recovered unitsSelected set sizeMeaning at the fixed conditions
353/5: the three reuse conditions are correctly recovered.
454/5: those same conditions plus the required-argument stop and return are correctly recovered.
555/5: those four units plus preservation of unrelated results are correctly recovered.

The other characteristics use their own ordinal quality meanings:

Characteristic3 means4 means5 means
OrderingRecoverabilityOrder is named, but wrong reconstruction remains likely.The intended order and relations needed for use are recoverable; consequential losses are clear.Conflicting order layers are handled and tested.
EventMechanismSupportEvents are coherent, but support strength is partly inferred from wording.Relation strength and reconstruction target are explicit.A wording or viewpoint change does not change recovered support strength.
ViewpointAgencyDisciplineViewpoint is useful, but agency or responsibility repair is incomplete.Viewpoint function and the source of a consequential attribution are recoverable.Reader can remove or swap viewpoint without losing source structure.
EngagementBoundednessInterest exists, but a needed source relation or condition for informed choice is missing.Interest supports the declared use while preserving the needed relations and conditions.A more engaging variant improves attention without lowering source recovery, with the comparison supported under the stated conditions.
SourceStructureSelectionFitSelected content partly serves the purpose, but a needed relation is omitted.Selected relations support the intended use and meaningful omissions are recoverable.A relevant changed case confirms the selection’s stated reach and limits.
LearningRouteReconstructionFitLearners retell the route and recover some source relations, but miss a relation needed for reliably applying it at a change of conditions.Learners reconstruct the essential relations and identify when to return to the source.Learners apply the route in a new case and identify the appropriate neighbouring method or responsible party.

NSTD.6:5.5 - Use an evaluation to improve a narrative

When repeated improvement is needed, use the narrative evaluation to identify the version, the intended change and the basis for comparing results. The following fields illustrate the input.

NarrativeRenderingImprovementLoopInput@Context:
  objectVersionRef: narrative rendering version, including a candidate
  evaluationResultRefs: selected `NSTD.6` rows
  improvementAim: raise one declared value without lowering protected trade-offs
  allowedChangedSlice: wording, source-return link, ordering marker, viewpoint repair, engagement device, generated source plan, or learning task
  protectedTradeoffSet: source fidelity, grounds for consequential claims, engagement, cost and reader burden
  expectedReEvaluationForm: reassess the affected NSTD.6 characteristics and any changed claim governed elsewhere

Regeneration can produce a candidate improvement when the versions, change, protected qualities and basis for reassessment remain identifiable. Another repair might add a source link and a task about the limit of an analogy. Compare the changed result under the selected evaluation in E.23. Increased engagement by itself establishes only the change in engagement, when that effect has been observed.

NSTD.6:5.6 - Evaluate a seminar handout or an architecture story

Use C.2.8 for structural amount in any narrative within its scope. NSTD.6 specifies the narrative, expression, observer and use conditions. Use C.33 when the question concerns the adequacy of an architecture description. Also assess source selection, access to the source, effort or usefulness if these affect the question.

In the seminar handout, a practitioner moves from the problem frame through forces and solution to checks. It names the source sections, explains the order, gives the learner a reconstruction task and specifies when to return to the source. Suppose the necessary relations are recoverable with the stated support. The other selected quality values may reach 4 or 5 under their own criteria. To quantify structural amount, use the structural scale.

A fluent architecture story explains that the chosen architecture felt coherent. It omits the rejected candidates, architectural characteristics and a route back to the sources. That defeats the use requiring those comparisons. Assign a structural amount only when the selection and conditions are established. High engagement can coexist with below-floor OrderingRecoverability or SourceReturnReadiness.

Reliability of a generator on unseen topics requires evidence about the generator across its stated range. A single generated narrative can still be inspected for a visible continuity error. If its source is unknown, leave fidelity and structural amount unassigned where their required basis is missing.

NSTD.6:6 - Bias-Annotation

A reader’s enjoyment, an expert’s approval, a fluency benchmark and performance on a test each establish different things. For this narrative, state the characteristics the use requires, their scales, thresholds, evidence, treatment of missing information and implications of a deficiency. Distinguish the characteristic being judged from the procedure used to evaluate it. Evidence, assurance and authority to admit, act or publish have their own criteria.

NSTD.6:7 - Conformance Checklist

CheckPassing condition
CC-NSTD6-1Evaluated object kind, declared use, and object-kind fit rule are explicit.
CC-NSTD6-2When developing or qualifying a specification, demonstrate that its criteria discriminate at least three cases: adequate use, a below-floor narrative and an object of another kind. Apply a qualified specification directly when it covers the present use.
CC-NSTD6-3Each characteristic has a declared qualitative comparison or a bound scale. Structural-amount values preserve their tuple, selection, conditions, granularity and evidence kind; generic quality labels do not substitute.
CC-NSTD6-4Value meanings, evidence basis, missingness rules, floor, exceptional meaning, and stop or reopen condition are declared.
CC-NSTD6-5Each row states the value or why it is unassigned and its evidence. For a deficiency, name the reason, any proposed repair and the conditions for reconsideration. Identify a responsible professional or decision holder when their contribution or authority matters.
CC-NSTD6-6Claims about measurement, evaluation, evidence, assurance, admission, decisions, publication or pattern quality have the grounds and, where applicable, the authority their use requires.
CC-NSTD6-7If repeated improvement is claimed, the E.22 or E.23 input names object version, NSTD.6 as evaluation, improvement aim, protected trade-offs, allowed change slice, cost and risk account, and expected re-evaluation form.
CC-NSTD6-8An improvement claim identifies a gain over an applicable baseline under the protected conditions. Evaluate the changed version or affected part through NSTD.6 or a specified stronger evaluation on a comparable basis.
CC-NSTD6-9Assess current quality from the product and its subject/use basis, fidelity from the source and resulting versions, and historical construction or repair from evidence of those actions. Address any admission claim under its own conditions.

NSTD.6:8 - Common Anti-Patterns and How to Avoid Them

Anti-patternWhat failsRepair
Fluency benchmark value as qualitySmoothness replaces structure recovery.Inspect the selected relations and evidence. Return a missing basis or repair a demonstrated loss through NSTD.1 through NSTD.3; fluency alone establishes neither a high nor a low amount.
Style repair as precision repairA nicer wording pass is treated as sufficient while relation kind, quality sense, language-state threshold, or coarsening loss remains hidden.Lower LanguageStatePrecisionAndCoarseningFit for the unresolved defect. Clarify it directly when its meaning is determined, or use the relevant precision, coarsening, explanation or language-state pattern for the remaining question. Re-evaluate the repaired rendering before claiming improvement.
Prompt loop as improvementThe worker keeps regenerating more engaging drafts without a named object version, allowed change slice, protected trade-offs, or re-evaluation.Open E.22 when needed, route the repair to E.23, and re-evaluate the changed version through NSTD.6; otherwise retain it as a candidate without claiming improvement.
Overextending a quality resultA wider evidence, assurance or admission claim lacks its own grounds.Use the quality result within its established scope. Use A.10 when a source-to-use account is implicit or must be reproducible, B.3 for a specified assurance claim and the applicable admission rule for a decision.
Evaluating a different objectA judgement about one narrative is used to rate its generator, source archive or whole course.Select the evaluation for that object. A script or generated draft can still be assessed as a narrative when it expresses one.
Reconstructed history as evidenceA plausible route inferred from a good product is reported as the author’s actual construction or repair.Retain the current-product result and mark the historical claim unestablished until its own evidence is obtained.
Treating an evaluation procedure as a characteristicA script used to obtain a value is put in the place of the property being judged.Specify the characteristic through A.19.ECS. Identify the automated measurement or evaluation procedure, the conditions in which it operates and the result it supplies.

NSTD.6:9 - Consequences

A reader can use the result to judge the version for the declared task. When improvement is wanted, the values, floors and observed losses identify changes that E.23 can compare. Establishing the grounds for reliance can require more time or access than a preliminary reading.

NSTD.6:10 - Rationale

Following A.19.ECS, identify the object and use, then select characteristics whose values distinguish adequate from deficient use. An improvement claim can be only as informative as the evaluation on which it depends.

NSTD.6:11 - SoTA-Echoing

FPF A.19.ECS and C.16 distinguish evaluation and measurement. C.2.8 supplies reader-relative structural amount and defines the conditions for its formal epiplexity estimate. C.33 addresses architecture-description adequacy. Effort, learning and usefulness can vary independently of amount. Use the language-state, coarsening or explanation method for the particular repair needed.

Castricato et al.’s model-theoretic account relates the information a narrator makes available to an evolving reader model. In Mengelkamp et al.’s reading-goal experiments, comprehension effects differed by condition and readers overestimated their performance. In Georgiou et al.’s narrative-memory study, recall after scrambled presentation could follow the original story order. Recall, recognition and reconstruction therefore warrant distinct questions. Use evidence of a reader’s task to assess what they recovered. The ordinal meanings above are this framework’s criteria.

Consider a reader judging a causal explanation. A fluent account can turn precedence into cause, so compare its causal relation with the source. If the question concerns what a reader understood, use their answer or an appropriate task when the available evidence warrants that work. For an expressive account, select the characteristics its purpose actually requires.

Ma et al.’s story-generation survey identifies continuing challenges in coherence, consistency, diversity, control and evaluation. Rahman et al.’s G-KMS uses structural, semantic and engine-execution checks for its RPG. Choose a check for the claim at issue. Inspect the draft for a visible continuity defect. Use appropriate evidence from a reader when the claim concerns their experience.

NSTD.6:12 - Relations

Uses A.19.ECS, A.17, A.18, C.2.8, C.16, C.16.Q, C.2.LS, A.16.1, A.16.2, A.6.3.CSC, E.17.EFP, E.10, A.6.P, C.33, E.22, E.23, B.4, A.10, B.3, E.17, C.35 and G.11. For an improvement comparison under E.23, identify the version, change, protected qualities, cost, risk and reassessment. Use B.4 for evolution under repeated use and G.11 when changed grounds affect reliance. Reconsider the affected judgement when its object, use, source, characteristics, language or precision conditions, cue, backoff, coarsening, explanation, structural basis, scale, threshold, evidence or proposed repair changes materially. The Architecture and Narrative Work Bridge relates structural recovery under C.2.8 to the separate architecture-adequacy question under C.33. The Semiotic and Language-Precision Bridge covers language, expression and precision, and the Source Use and Refresh Map covers support and currency. Consult NSTD.Reference:11 for the meanings and sources of the characteristic names.

NSTD.6:End

Referenced in the corpus

71 literal mentions in other sections. Read their context to establish the relation.