Instructional-material evaluation profile
A material designer needs to know whether an explanation, worksheet, course route, or instructional interface supports the work promised to its audience, and what to change when it does not. The instructional-material profile selects that bounded use from HCD. The object being judged is the material’s contribution under stated task and assistance conditions; human performance, the teacher service, and later learning remain separately evidenced questions.
Begin with the reader work, the audience’s prerequisites, the available material and help, and the decision that will use the judgement. If the evaluation specification is absent or unsuitable, HCD.18 constructs it. If a compatible specification is already available, HCD.19 applies it directly to the material and representative reader work. The useful return is a qualified material value or evidence gap, the limiting defect and repair reach, and the affected use that can test a proposed change.
This is a declarative bounded-use profile of the HCD repertoire. It does not make all HCD patterns stages of material evaluation or assert Method parthood. HCD.3/.6 can supply a target and task; HCD.7/.9 can supply or expose the assistance and practice contribution. Qualified equivalent inputs permit direct entry. A personal development recommendation still needs the person’s evidence. A reference framework used for a professional decision retains its own evaluation; it becomes an instructional-material use only when instruction is promised.
Choose properties from the promised work
Recover the task denominator from the promised work and qualified subject requirements, including prerequisites, consequential variations, and critical errors. Then compare it with the actual explanations, examples, practice, and usable returns. The present headings can help locate content but cannot remove a necessary task from that denominator.
The following families are selectable domain content, not a fixed vector. Keep a characteristic only when its values can change use, repair, evidence collection, or protection of another useful contribution. HCD.18 makes the selected set complete for the particular decision.
| Characteristic family | What distinguishes useful support from a limiting defect | Typical repair and evidence |
|---|---|---|
| Task and dependency coverage | Every promised task and consequential variation has its needed operations and prerequisites available. | Trace required work to actual contributions. Supply an absent operation, usable qualified return, or narrower promise. |
| Subject validity and distinction integrity | Claims, examples, diagrams, and criteria preserve domain-correct kinds, relations, units, conditions, and justified claim strength. | Test a plausible countercase against a qualified subject basis. Correct the relation and its actual dependents. |
| Extractable structural information, when the amount matters | What selected structure this reader can correctly extract from the material episteme through this expression under the stated conditions, using C.2.8. | Name recovered and missing relations; use fixed semantic units if counting. Repair the missing or false relation, expression or usable return. Keep amount, effort, material adequacy and learning distinct. |
| Explanatory sufficiency | The reader can recover why the next move follows and when it changes, with the declared assistance. | Follow the explanation through a worked case to use. Restore a missing ground, contrast, transition, or operation; obtain relevant reader evidence when the claim needs it. |
| Example discrimination | The example makes a consequential difference between the intended rule and a plausible wrong rule. | Try both rules on the examples. Add or change a contrast where their predictions differ. |
| Task and criterion validity | The task elicits the promised contribution, and the criterion accepts valid alternatives while rejecting a plausible error. | Independently solve and challenge the task and key. Repair the criterion before blaming the reader or changing a correct explanation. |
| Feedback and continuation support | The arrangement lets a reader identify a discrepancy, correct it, and undertake a relevant next attempt. | Inspect and observe the feedback’s function. Supply criterion-bearing help and a retry, or repair an unavailable promised service at its owner. |
| Concrete and expressive support | An action, observation, example, image, or memorable cue reveals the relation needed in use, with its limits recoverable. | Retain or restore decision-bearing detail and vivid examples. A quiet reference needs no decoration merely to appear engaging. |
| Representation semantic fit | The chosen text, image, table, or interaction exposes the needed relation accurately and accessibly. | Repair the mapping, label, relation, pacing, or relevant alternative representation. Inspect an export separately when its rendering changes access. |
| Language comprehensibility for the audience | The reader can interpret the claim and scope without an accidental negation, ambiguous referent, or unintroduced language burden. | Repair the meaning-bearing span in the declared language. A local authorial-voice requirement remains local. |
| Currentness of relied-on content | The particular claim and dependent operation fit the declared subject or source state. | Recheck the changed premise and its consumers. Age alone does not decide whether a claim remains usable. |
| Practical addressability and access | The reader can reach the source, prerequisite, task, or feedback needed at this point. | Repair the actual target or accessible return. A broken link is not evidence that the underlying subject claim is false. |
| Narrative attachment, when live | The current question, new contribution, reason for placement, and later payoff connect across the actual sequence. | Use NSTD.8 to inspect both endpoints and restore the needed relation or placement; no connective-word quota is implied. |
| Progressive reconstruction support, when live | The required operation fits the structure already introduced or legitimately available, with adequate support for the new relation. | Use NSTD.8 to supply a contrast, map, worked relation, recap, usable return, or better order while retaining useful difficulty. |
These properties can share observations without sharing a value. A correct explanation can coexist with an invalid key. A teacher can rescue an explanation while the example still fails to distinguish a rival rule. A diagram’s semantic fit can be adequate while a particular export makes its labels unreadable. Split questions when their evidence or repair differs; do not count one shared observation several times as independent evidence.
Interpret material value separately from evidence and fit
For material-use and repair questions, the default is a qualified ordinal judgement. A numeric 0–5 rendering is optional. A selected C.2.8 structural amount instead uses its own qualitative comparison, count, fraction or other justified structural scale; the following material-quality labels neither measure that amount nor set its threshold. When using the 0–5 material rendering, bind the following meanings to the exact property, material grain, task family, and support arrangement before applying them.
| Value | Material meaning for the declared use |
|---|---|
| 0 | The material demonstrably supplies no usable support for the required property, or defeats an explicitly declared non-compensable condition at this evaluation grain. |
| 1 | Some relevant support exists, but a material part of the promised work remains out of reach. |
| 2 | Useful work is recoverable, but repair must rebuild a connected central explanation–task chain or several consequential dependent uses. |
| 3 | The main work is recoverable; a bounded correction to the identified relation, task, representation, or return is still needed. |
| 4 | The property is adequate for the declared tasks, audience, and available support on the evidence required by the specification; no material defect is currently established for that use. |
| 5 | The property has defensible additional useful resilience across the specification’s consequential variations or boundary cases; the advantage and its evidence over 4 are stated. |
The numbers denote ordered meanings, not equal intervals. A severe local defect does not automatically make a whole book 0. Trace its effect on the declared task family and any actual non-compensable condition. A high average cannot compensate a failed condition that the use cannot trade away.
A floor belongs to the consuming decision, not to HCD as a whole. The Time-use S1 example in HCD.18 chooses 4 for each selected property before relying on its bounded supported introduction. That is a local rule. A diagnostic, controlled probe, or different consequence can require a different decision without changing what a value means.
Evidence accompanies the value. Source inspection, an actual reader attempt, conditional inference, disputed criterion, and an unobserved contribution are different grounds. Wrong object kind is a fit decision before invocation. A required coordinate that cannot be judged returns its explicit evidence or criterion defect and the conclusion it blocks. It does not disappear from a complete result and does not receive 0 for ignorance.
For an assigned value, explain why adjacent alternatives would understate or overstate the evidence. Distinguish a central connected repair from a bounded one by the changes needed in use, not error count. A value of 5 requires useful resilience beyond 4; actual manufacturing or repair history is additional evidence only when a historical claim is live. A value of 4 can still have a worthwhile improvement. Compare that improvement’s benefit with its cost and protected contributions.
Earlier evaluations under other specifications retain their original value meanings. For example, a local Guide scale whose 0 includes wrong kind or no basis is not numerically equivalent to this material scale. A comparison may relate “connected repair” to “bounded repair” while explicitly declining numeric conversion.
Use the two methods together without adding a lifecycle
HCD.18 supplies a specification when one is needed. Its complete result includes fit, task denominator, selected properties and scale meanings, evidence and missingness, result interpretation, contrasts, protected trade-offs, and stops or reopens. It inherits general specification construction from A.19.ECS and supplies the instructional domain choices.
HCD.19 uses that specification to connect the exact material, subject criterion, available reader responses, actual assistance and repair. Compatible observations and qualified source inspection can finish a bounded diagnosis; missing evidence still withholds its dependent claim. Select further reader work only when its obtainable contribution warrants the full burden under the actual access, participation and protection conditions. Its result feeds E.23 improvement. If selecting among alternatives is live, E.22 supplies comparison under a common qualified basis. A new book format or a different teacher arrangement may change that basis; preserve earlier results instead of quietly relabelling them.
For long material, a whole-route conclusion needs connected inspection of the complete declared route and its cross-part dependencies. For a local defect, inspect the smallest complete semantic reach and reuse unaffected evidence. A moved explanation can change the next-use prefix even if its words are unchanged.
NSTD contributes only when the narrative question is live. NSTD.6 distinguishes the current product, promised source-to-result fidelity, and actual construction history. NSTD.8 supplies the attachment and available-prefix method. NarrativeRenderingEpiplexity is NSTD.6’s narrative specialization of C.2.8 over the expressed episteme, expression and observer, qualified by the selected structure and reading conditions. Reuse a structural result only when those values match. It is not observed cognitive load or learning effectiveness, and non-narrative material can use C.2.8 directly through HCD.18/.19. Retain a human report of difficulty and narrow any untested audience or experienced-burden claim now; the supported material diagnosis can finish with that human question unresolved. Select an appropriately timed observation of the actual reader only when its obtainable contribution can change the receiving decision enough to warrant the full design, preparation, reader, provider, interpretation, delay and displaced-work burden, with the needed access, participation and protection. Selecting the observation does not establish its result.
The time-accounting case makes the joint result concrete. M0’s uninterrupted example fits both active effort and elapsed turnaround. A prepared reader correctly separates them anyway and identifies its prior-knowledge contribution. M1 repairs the definitions, columns, interruption example, and population denominator. The same reader then correctly compares three consumed person-hours with four available person-hours in a changed two-person case. The material repair and bounded changed-condition use are supported; novice learning is not observed.
HCD.18’s full Time-use S1 specification also requires a concurrent-person example, an inspected task-and-key arrangement, and audience-compatible explanatory and feedback evidence. HCD.19’s complete replay returns those residual repairs or gaps instead of treating the successful agent answers as a complete positive evaluation. The practitioner can now decide exactly what to add and what to observe next.
Why this profile is useful, and when a simpler alternative suffices
A subject expert’s inspection is sufficient for a narrow false claim or unit relation. It is not enough for an audience-dependent claim that a reader can recover the explanation. A conventional task-and-scoring instrument may already supply an adequate specification; use it directly when its task, population, assistance, and inference fit. A format-specific production check is appropriate for an export defect, while a material-semantic defect can cross several exports.
Compared with a checklist of headings or visible features, this profile costs task and contrast construction but exposes missing work and distinguishes repairs. Compared with one post-reading score, it preserves the material and help that made the response possible. Compared with demanding a full learner study for every edit, it permits a qualified source diagnosis and bounded repair while naming the stronger evidence still needed.
Its principal risk is overinterpretation: an expert can silently repair a source, an evaluator can teach through a cue, and a fluent account can conceal a missing relation. Preserve first responses, inspect subject criteria and material contribution, and qualify the result by actual audience and support. Reopen only the properties and uses changed by a new premise, material change, meaningful reader result, or recurrent scale disagreement.
Source-informed choices and limits
The best-known line used here is task- and interpretation-sensitive evaluation, with examples that distinguish rival rules and observations that preserve assistance and continuation. HCD.18 and HCD.19 carry the comparisons, worked consequences, and source-specific reopens. These are professional adaptations; the studies do not validate this profile or its local ordinal anchors.
Wesenberg et al. (2025) supply failure evidence for ambiguous worked examples in bounded topics. McCrudden et al. (2022) and 2024 inform the required relation between principles, examples, and self-explanation. The resulting choice is to make the relation consequential in a task; it is not a universal ordering rule.
Noetel et al. (2022) inform mechanism-specific representation repair. Sinha and Kapur (2021) support distinguishing preparation followed by resolving instruction from unsupported difficulty. Wisniewski et al. (2020) support testing feedback’s function rather than its presence. None supplies a universal local threshold, fixed reader count, or whole-book adequacy finding.
Bastani et al. (2025) inform the separation of assisted and later independent performance in their studied arrangement. Krieglstein et al. (2025) discipline the timing and scope of experienced-burden observations. Bavaresco et al. (2025) supply task-dependent limits on substituting LLM judgements for human ones. Their practical consequence is qualified evidence, not simulated human learning or a ban on useful support.