# A10 instruments and governance records · version 1.0

These are original proposed templates, not validated scales or executed permissions. No student, family or cultural contribution is included. Final instruments and translated forms must be locked after an independent pilot and before allocation.

## 1. Proposed primary belonging instrument

Proposed measure: Carol Goodenow's 18-item Psychological Sense of School Membership (1993), Psychology in the Schools, 30(1),79–90. DOI:10.1002/1520-6807(199301)30:1<79::AID-PITS2310300113>3.0.CO;2-X.

The copyrighted item text is not reproduced here. Before recruitment obtain the applicable permission, original wording and scoring key; verify any rights to translate or adapt. Do not substitute a reconstructed item list from an AI response. The proposal uses the 1–5 item response metric and the original reverse-key directions; an occasion score is the mean of the keyed responses when at least16 of18 are present. This ≥16 completion/pro-rating rule is a proposed study rule, not claimed as a validated rule from Goodenow. If fewer than16 are answered, the occasion score is missing, not zero. Record item missingness separately. No response is graded.

Independent pilot: cognitive interviews across intended language and accessibility routes; burden and refusals; ordinal factor and threshold behavior across time/language; feasibility of the proposed completion rule. The pilot is not part of the confirmatory cohort. Any replacement primary measure or revised key requires a versioned amendment before allocation. The measure does not assess cultural authenticity, political agreement, immigration, religion or willingness to publish.

## 2. Original disciplinary-transfer task blueprint

Forms A/B are counterbalanced within schools at baseline/week24. Form C introduces a new setting at week36. Task forms are fixed and trial access restricted before allocation. All forms supply enough evidence to reason without personal disclosure or privileged local knowledge. Stimulus examples are fictional and require independent curriculum review, permissions and accessibility checks.

A candidate task presents two explanations for a change in a public place, such as a fictional neighborhood's heat exposure. Evidence cards include measurements, uncertainty, a map with an incomplete variable and two stakeholders' competing interpretations. Ask the learner to explain the pattern, select useful evidence, identify an omitted constraint and revise a recommendation after receiving new data. A parallel form uses a fictional water-use question; the delayed form changes the mechanism and decision constraints. Pilot matching must establish comparable demand rather than assume that related topics are equivalent.

### Proposed 0–16 rubric (not a validated scale)

Score four dimensions0–4. Raters assess reasoning, not agreement with the school or quantity of cultural detail. No identity disclosure adds points.

| Dimension |0|1|2|3|4|
|---|---|---|---|---|---|
| Disciplinary explanation |No interpretable explanation|Names a factor without relation|Plausible relation but major mechanism missing|Coherent disciplinary mechanism with a stated limit|Coherent mechanism distinguishes a plausible rival and explains its limit|
| Evidence discrimination |No evidence used|Repeats a card without relevance|Selects relevant evidence but ignores a major conflict|Weighs relevant and conflicting evidence with a reason|Also identifies an uncertainty or missing comparison that could alter the judgment|
| Reasoned revision |No revision or rationale|Changes answer without reason|Connects a change to one new fact|Explains what changed and what remained defensible|Revises the model/decision proportionately and states what would reverse it|
| Transfer to changed constraint |No response|Repeats the original answer unchanged|Adapts one surface feature|Applies the mechanism with an appropriate changed constraint|Explains where the prior mechanism transfers and where it does not|

Two independent blinded human raters score every artifact; primary task score is their mean total. Keep initial scores, timing, uncertainty and form version. A third rater audits a random10% plus designated reliability concerns; those audit scores do not selectively overwrite the primary mean. Report an absolute-agreement reliability statistic, score distribution and disagreements; calibration uses separate pilot work. Remove condition labels, school/place names and obvious intervention slogans before assignment to raters where this preserves substantive meaning. Raters guess condition only after scoring.

## 3. Representation and decision charter

For each cycle record: disciplinary goal; nonpersonal task route; source/material/version; who owns or controls proposed material; permission scope (instruction, research, recording, quotation, public display, future reuse separately); any community authority actually delegated; individual consent/assent route; accessible challenge route; responsible decision-holder; response deadline; appeal route; limits and unresolved disagreement; final response and version; retention/deletion rule.

No blank entry defaults to permission. An individual may consent only within their authority. A contributor does not speak for every member of a reported demographic group. Collective decisions cannot compel a student's personal disclosure. Where a legitimate rights-holder refuses representation/reuse, use an authorized alternative; do not resolve the refusal by paraphrasing protected knowledge through AI.

## 4. Governance-process fidelity (descriptive indicators)

For each six-week cycle: Was there an inspectable representation record? Could participants access it in a usable language/mode? Were decision rights explained before contribution? How many challenges arose, received a reasoned answer within the agreed period and changed a design? How many remained disputed? Was refusal/private participation usable without penalty? Report the denominator and opportunity, not a raw rank of schools. A school with fewer reported objections may be safer, more satisfied or less safe for dissent; the count alone does not discriminate among those explanations.

## 5. Burden and exposure log

Fortnightly teacher log: scheduled paid design minutes, actual paid design minutes, unpaid additional minutes, replaced activity, facilitator minutes, accessibility support, technology/material expenditure, and unusual events. Contributor log: compensated time, payment received, preferred/private mode, barriers and permission changes. Student optional check-in: time burden, unwanted disclosure pressure, availability of a private route and whether asking for change felt consequential. These original prompts are descriptive, not a validated culture or belonging scale.

At school level record shared facilitators/design teams and materials actually exchanged with other participating schools; do not collect a personal student social graph. Report baseline network concerns before allocation and subsequent contact separately. Avoid soliciting identifying stories in an exposure log.

## 6. Optional qualitative prompts

What could you change in the task? Describe a decision without naming someone who has not consented. Was there a time you disagreed but chose not to use the formal route? What made the response useful or unhelpful? Did a private/fictitious setting let you participate fully? What additional work did the process create? What would you stop? Participants may skip every question. Do not ask someone to prove cultural authenticity or identify who 'really' represents a community.

A disagreement log records competing interpretations, evidence, authority, reasons and resolution status. Public summaries remove small-group identifiers and sensitive quotations. Member reflection checks whether an account is fair; it is not a required declaration of consensus or a veto that permits an institution to erase unfavorable findings.