ICEQC-R-2020-08 — Unequal Effects of Disruption by Disability, Gender, Location and Income

Exposure, response and outcome are distinct defines a particular comparison within the effect definition. For exposure, response and outcome are distinct, the affected population is learners experiencing varied disruption and institutional response, and the immediate evidence concerns closure, service received and later condition. In comparing exposure, response and outcome are distinct, a valid account states who was exposed, what institutional response was offered, who could use it and which learner condition was later observed. Within evidence on exposure, response and outcome are distinct, these stages should not collapse into a single indicator.[REF-05] [REF-38] [REF-52] [REF-56]

The principal risk is that a common national closure period is treated as equal learner exposure. In comparing exposure, response and outcome are distinct, that error can make excluded learners invisible or assign unequal effects to personal characteristics rather than institutions and material conditions. Within evidence on exposure, response and outcome are distinct, review should reconstruct entry into each source, actual service receipt and the barrier controlled by households, schools or public authorities.[REF-05] [REF-38] [REF-52]

The recommended response is to report each stage separately. Within evidence on exposure, response and outcome are distinct, reporting should include counts, group levels, reference period, source and material missingness. For interpretation of exposure, response and outcome are distinct, a gap or ratio should appear beside absolute conditions so apparent parity does not conceal low provision for all. In decisions about exposure, response and outcome are distinct, where the denominator or category is unstable, the conclusion should remain bounded.[REF-38] [REF-52] [REF-56]

Evidence for exposure, response and outcome are distinct should distinguish administrative status, household access, actual contact, educational use and later outcome. For interpretation of exposure, response and outcome are distinct, surveys, registers, provider records and experience evidence observe different populations. In decisions about exposure, response and outcome are distinct, agreement can strengthen a finding, while disagreement may reveal excluded settings, timing or definition differences that require explanation.[REF-05] [REF-56]

Intersectional interpretation of exposure, response and outcome are distinct should select characteristics tied to a plausible barrier and public decision. In decisions about exposure, response and outcome are distinct, disability, gender, location and income can combine with age, language or migration. For correction of exposure, response and outcome are distinct, counts, intervals and disclosure risk should determine the precision of publication. For exposure, response and outcome are distinct, small groups require neither invisibility nor unwarranted certainty.[REF-38] [REF-52] [REF-56]

Missing evidence for exposure, response and outcome are distinct is itself potentially unequal. For correction of exposure, response and outcome are distinct, learners unable to access education media may also be unable to answer connected surveys or submit work. For exposure, response and outcome are distinct, non-response should therefore be investigated by route and group. In comparing exposure, response and outcome are distinct, imputation or complete-case analysis should state assumptions and should not erase the likely direction of exclusion.[REF-05] [REF-38] [REF-52] [REF-56]

Causal interpretation of exposure, response and outcome are distinct should preserve baseline inequality, additional disruption, response differences and later change. For exposure, response and outcome are distinct, a post-disruption gap does not by itself identify its cause. In comparing exposure, response and outcome are distinct, evidence of exposure, timing and alternative explanations supports a bounded contribution claim; otherwise the report should describe association and the practical barrier requiring correction.[REF-38] [REF-52] [REF-56]

Public accountability for exposure, response and outcome are distinct should name the competent owner, resource, accessible correction and review date. In comparing exposure, response and outcome are distinct, learners need safe complaint and continuity routes. Within evidence on exposure, response and outcome are distinct, later review should ask whether the corrected service reached the group and whether participation, learning or wellbeing improved, while retaining the original definition and evidence limitations.[REF-05] [REF-38] [REF-52] [REF-56]

Availability, access, receipt and use defines a particular comparison within the service chain. For availability, access, receipt and use, the affected population is learners offered remote arrangements, and the immediate evidence concerns offer, usable access, actual receipt and educational use. In comparing availability, access, receipt and use, a valid account states who was exposed, what institutional response was offered, who could use it and which learner condition was later observed. Within evidence on availability, access, receipt and use, these stages should not collapse into a single indicator.[REF-07] [REF-49] [REF-55] [REF-56]

The principal risk is that resources announced are counted as equitable education. In comparing availability, access, receipt and use, that error can make excluded learners invisible or assign unequal effects to personal characteristics rather than institutions and material conditions. Within evidence on availability, access, receipt and use, review should reconstruct entry into each source, actual service receipt and the barrier controlled by households, schools or public authorities.[REF-07] [REF-49] [REF-55]

The recommended response is to trace the chain from policy to learner use. Within evidence on availability, access, receipt and use, reporting should include counts, group levels, reference period, source and material missingness. For interpretation of availability, access, receipt and use, a gap or ratio should appear beside absolute conditions so apparent parity does not conceal low provision for all. In decisions about availability, access, receipt and use, where the denominator or category is unstable, the conclusion should remain bounded.[REF-49] [REF-55] [REF-56]

Evidence for availability, access, receipt and use should distinguish administrative status, household access, actual contact, educational use and later outcome. For interpretation of availability, access, receipt and use, surveys, registers, provider records and experience evidence observe different populations. In decisions about availability, access, receipt and use, agreement can strengthen a finding, while disagreement may reveal excluded settings, timing or definition differences that require explanation.[REF-07] [REF-56]

Intersectional interpretation of availability, access, receipt and use should select characteristics tied to a plausible barrier and public decision. In decisions about availability, access, receipt and use, disability, gender, location and income can combine with age, language or migration. For correction of availability, access, receipt and use, counts, intervals and disclosure risk should determine the precision of publication. For availability, access, receipt and use, small groups require neither invisibility nor unwarranted certainty.[REF-49] [REF-55] [REF-56]

Missing evidence for availability, access, receipt and use is itself potentially unequal. For correction of availability, access, receipt and use, learners unable to access education media may also be unable to answer connected surveys or submit work. For availability, access, receipt and use, non-response should therefore be investigated by route and group. In comparing availability, access, receipt and use, imputation or complete-case analysis should state assumptions and should not erase the likely direction of exclusion.[REF-07] [REF-49] [REF-55] [REF-56]

Causal interpretation of availability, access, receipt and use should preserve baseline inequality, additional disruption, response differences and later change. For availability, access, receipt and use, a post-disruption gap does not by itself identify its cause. In comparing availability, access, receipt and use, evidence of exposure, timing and alternative explanations supports a bounded contribution claim; otherwise the report should describe association and the practical barrier requiring correction.[REF-49] [REF-55] [REF-56]

Public accountability for availability, access, receipt and use should name the competent owner, resource, accessible correction and review date. In comparing availability, access, receipt and use, learners need safe complaint and continuity routes. Within evidence on availability, access, receipt and use, later review should ask whether the corrected service reached the group and whether participation, learning or wellbeing improved, while retaining the original definition and evidence limitations.[REF-07] [REF-49] [REF-55] [REF-56]

Participation, learning and wellbeing defines a particular comparison within the outcome plurality. For participation, learning and wellbeing, the affected population is learners affected across several education conditions, and the immediate evidence concerns contact, attendance, learning, safety and wellbeing. In comparing participation, learning and wellbeing, a valid account states who was exposed, what institutional response was offered, who could use it and which learner condition was later observed. Within evidence on participation, learning and wellbeing, these stages should not collapse into a single indicator.[REF-10] [REF-12] [REF-50] [REF-53]

The principal risk is that one task or score stands for the whole disruption effect. In comparing participation, learning and wellbeing, that error can make excluded learners invisible or assign unequal effects to personal characteristics rather than institutions and material conditions. Within evidence on participation, learning and wellbeing, review should reconstruct entry into each source, actual service receipt and the barrier controlled by households, schools or public authorities.[REF-10] [REF-12] [REF-50]

The recommended response is to retain distinct outcome domains. Within evidence on participation, learning and wellbeing, reporting should include counts, group levels, reference period, source and material missingness. For interpretation of participation, learning and wellbeing, a gap or ratio should appear beside absolute conditions so apparent parity does not conceal low provision for all. In decisions about participation, learning and wellbeing, where the denominator or category is unstable, the conclusion should remain bounded.[REF-12] [REF-50] [REF-53]

Evidence for participation, learning and wellbeing should distinguish administrative status, household access, actual contact, educational use and later outcome. For interpretation of participation, learning and wellbeing, surveys, registers, provider records and experience evidence observe different populations. In decisions about participation, learning and wellbeing, agreement can strengthen a finding, while disagreement may reveal excluded settings, timing or definition differences that require explanation.[REF-10] [REF-53]

Intersectional interpretation of participation, learning and wellbeing should select characteristics tied to a plausible barrier and public decision. In decisions about participation, learning and wellbeing, disability, gender, location and income can combine with age, language or migration. For correction of participation, learning and wellbeing, counts, intervals and disclosure risk should determine the precision of publication. For participation, learning and wellbeing, small groups require neither invisibility nor unwarranted certainty.[REF-12] [REF-50] [REF-53]

Missing evidence for participation, learning and wellbeing is itself potentially unequal. For correction of participation, learning and wellbeing, learners unable to access education media may also be unable to answer connected surveys or submit work. For participation, learning and wellbeing, non-response should therefore be investigated by route and group. In comparing participation, learning and wellbeing, imputation or complete-case analysis should state assumptions and should not erase the likely direction of exclusion.[REF-10] [REF-12] [REF-50] [REF-53]

Causal interpretation of participation, learning and wellbeing should preserve baseline inequality, additional disruption, response differences and later change. For participation, learning and wellbeing, a post-disruption gap does not by itself identify its cause. In comparing participation, learning and wellbeing, evidence of exposure, timing and alternative explanations supports a bounded contribution claim; otherwise the report should describe association and the practical barrier requiring correction.[REF-12] [REF-50] [REF-53]

Public accountability for participation, learning and wellbeing should name the competent owner, resource, accessible correction and review date. In comparing participation, learning and wellbeing, learners need safe complaint and continuity routes. Within evidence on participation, learning and wellbeing, later review should ask whether the corrected service reached the group and whether participation, learning or wellbeing improved, while retaining the original definition and evidence limitations.[REF-10] [REF-12] [REF-50] [REF-53]

Absolute levels and group gaps defines a particular comparison within the inequality account. For absolute levels and group gaps, the affected population is groups with different starting levels and later change, and the immediate evidence concerns group level, difference and change. In comparing absolute levels and group gaps, a valid account states who was exposed, what institutional response was offered, who could use it and which learner condition was later observed. Within evidence on absolute levels and group gaps, these stages should not collapse into a single indicator.[REF-03] [REF-10] [REF-38] [REF-44]

The principal risk is that a narrowing gap hides deterioration for every group. In comparing absolute levels and group gaps, that error can make excluded learners invisible or assign unequal effects to personal characteristics rather than institutions and material conditions. Within evidence on absolute levels and group gaps, review should reconstruct entry into each source, actual service receipt and the barrier controlled by households, schools or public authorities.[REF-03] [REF-10] [REF-38]

The recommended response is to publish absolute conditions beside disparities. Within evidence on absolute levels and group gaps, reporting should include counts, group levels, reference period, source and material missingness. For interpretation of absolute levels and group gaps, a gap or ratio should appear beside absolute conditions so apparent parity does not conceal low provision for all. In decisions about absolute levels and group gaps, where the denominator or category is unstable, the conclusion should remain bounded.[REF-10] [REF-38] [REF-44]

Evidence for absolute levels and group gaps should distinguish administrative status, household access, actual contact, educational use and later outcome. For interpretation of absolute levels and group gaps, surveys, registers, provider records and experience evidence observe different populations. In decisions about absolute levels and group gaps, agreement can strengthen a finding, while disagreement may reveal excluded settings, timing or definition differences that require explanation.[REF-03] [REF-44]

Intersectional interpretation of absolute levels and group gaps should select characteristics tied to a plausible barrier and public decision. In decisions about absolute levels and group gaps, disability, gender, location and income can combine with age, language or migration. For correction of absolute levels and group gaps, counts, intervals and disclosure risk should determine the precision of publication. For absolute levels and group gaps, small groups require neither invisibility nor unwarranted certainty.[REF-10] [REF-38] [REF-44]

Missing evidence for absolute levels and group gaps is itself potentially unequal. For correction of absolute levels and group gaps, learners unable to access education media may also be unable to answer connected surveys or submit work. For absolute levels and group gaps, non-response should therefore be investigated by route and group. In comparing absolute levels and group gaps, imputation or complete-case analysis should state assumptions and should not erase the likely direction of exclusion.[REF-03] [REF-10] [REF-38] [REF-44]

Causal interpretation of absolute levels and group gaps should preserve baseline inequality, additional disruption, response differences and later change. For absolute levels and group gaps, a post-disruption gap does not by itself identify its cause. In comparing absolute levels and group gaps, evidence of exposure, timing and alternative explanations supports a bounded contribution claim; otherwise the report should describe association and the practical barrier requiring correction.[REF-10] [REF-38] [REF-44]

Public accountability for absolute levels and group gaps should name the competent owner, resource, accessible correction and review date. In comparing absolute levels and group gaps, learners need safe complaint and continuity routes. Within evidence on absolute levels and group gaps, later review should ask whether the corrected service reached the group and whether participation, learning or wellbeing improved, while retaining the original definition and evidence limitations.[REF-03] [REF-10] [REF-38] [REF-44]

Starting inequality and additional disruption defines a particular comparison within the baseline relation. For starting inequality and additional disruption, the affected population is learners entering disruption with unequal opportunity, and the immediate evidence concerns pre-disruption level and additional change. In comparing starting inequality and additional disruption, a valid account states who was exposed, what institutional response was offered, who could use it and which learner condition was later observed. Within evidence on starting inequality and additional disruption, these stages should not collapse into a single indicator.[REF-04] [REF-29] [REF-38] [REF-53]

The principal risk is that all observed inequality is attributed to disruption or prior inequality is ignored. In comparing starting inequality and additional disruption, that error can make excluded learners invisible or assign unequal effects to personal characteristics rather than institutions and material conditions. Within evidence on starting inequality and additional disruption, review should reconstruct entry into each source, actual service receipt and the barrier controlled by households, schools or public authorities.[REF-04] [REF-29] [REF-38]

The recommended response is to separate baseline, exposure and change where evidence permits. Within evidence on starting inequality and additional disruption, reporting should include counts, group levels, reference period, source and material missingness. For interpretation of starting inequality and additional disruption, a gap or ratio should appear beside absolute conditions so apparent parity does not conceal low provision for all. In decisions about starting inequality and additional disruption, where the denominator or category is unstable, the conclusion should remain bounded.[REF-29] [REF-38] [REF-53]

Evidence for starting inequality and additional disruption should distinguish administrative status, household access, actual contact, educational use and later outcome. For interpretation of starting inequality and additional disruption, surveys, registers, provider records and experience evidence observe different populations. In decisions about starting inequality and additional disruption, agreement can strengthen a finding, while disagreement may reveal excluded settings, timing or definition differences that require explanation.[REF-04] [REF-53]

Intersectional interpretation of starting inequality and additional disruption should select characteristics tied to a plausible barrier and public decision. In decisions about starting inequality and additional disruption, disability, gender, location and income can combine with age, language or migration. For correction of starting inequality and additional disruption, counts, intervals and disclosure risk should determine the precision of publication. For starting inequality and additional disruption, small groups require neither invisibility nor unwarranted certainty.[REF-29] [REF-38] [REF-53]

Missing evidence for starting inequality and additional disruption is itself potentially unequal. For correction of starting inequality and additional disruption, learners unable to access education media may also be unable to answer connected surveys or submit work. For starting inequality and additional disruption, non-response should therefore be investigated by route and group. In comparing starting inequality and additional disruption, imputation or complete-case analysis should state assumptions and should not erase the likely direction of exclusion.[REF-04] [REF-29] [REF-38] [REF-53]

Causal interpretation of starting inequality and additional disruption should preserve baseline inequality, additional disruption, response differences and later change. For starting inequality and additional disruption, a post-disruption gap does not by itself identify its cause. In comparing starting inequality and additional disruption, evidence of exposure, timing and alternative explanations supports a bounded contribution claim; otherwise the report should describe association and the practical barrier requiring correction.[REF-29] [REF-38] [REF-53]

Public accountability for starting inequality and additional disruption should name the competent owner, resource, accessible correction and review date. In comparing starting inequality and additional disruption, learners need safe complaint and continuity routes. Within evidence on starting inequality and additional disruption, later review should ask whether the corrected service reached the group and whether participation, learning or wellbeing improved, while retaining the original definition and evidence limitations.[REF-04] [REF-29] [REF-38] [REF-53]

Causal restraint under changing conditions defines a particular comparison within the interpretive boundary. For causal restraint under changing conditions, the affected population is authorities comparing rapidly changing populations and services, and the immediate evidence concerns timing, exposure and plausible alternatives. In comparing causal restraint under changing conditions, a valid account states who was exposed, what institutional response was offered, who could use it and which learner condition was later observed. Within evidence on causal restraint under changing conditions, these stages should not collapse into a single indicator.[REF-09] [REF-19] [REF-38] [REF-52]

The principal risk is that post-disruption differences are automatically attributed to one cause. In comparing causal restraint under changing conditions, that error can make excluded learners invisible or assign unequal effects to personal characteristics rather than institutions and material conditions. Within evidence on causal restraint under changing conditions, review should reconstruct entry into each source, actual service receipt and the barrier controlled by households, schools or public authorities.[REF-09] [REF-19] [REF-38]

The recommended response is to use bounded contribution claims and state uncertainty. Within evidence on causal restraint under changing conditions, reporting should include counts, group levels, reference period, source and material missingness. For interpretation of causal restraint under changing conditions, a gap or ratio should appear beside absolute conditions so apparent parity does not conceal low provision for all. In decisions about causal restraint under changing conditions, where the denominator or category is unstable, the conclusion should remain bounded.[REF-19] [REF-38] [REF-52]

Evidence for causal restraint under changing conditions should distinguish administrative status, household access, actual contact, educational use and later outcome. For interpretation of causal restraint under changing conditions, surveys, registers, provider records and experience evidence observe different populations. In decisions about causal restraint under changing conditions, agreement can strengthen a finding, while disagreement may reveal excluded settings, timing or definition differences that require explanation.[REF-09] [REF-52]

Intersectional interpretation of causal restraint under changing conditions should select characteristics tied to a plausible barrier and public decision. In decisions about causal restraint under changing conditions, disability, gender, location and income can combine with age, language or migration. For correction of causal restraint under changing conditions, counts, intervals and disclosure risk should determine the precision of publication. For causal restraint under changing conditions, small groups require neither invisibility nor unwarranted certainty.[REF-19] [REF-38] [REF-52]

Missing evidence for causal restraint under changing conditions is itself potentially unequal. For correction of causal restraint under changing conditions, learners unable to access education media may also be unable to answer connected surveys or submit work. For causal restraint under changing conditions, non-response should therefore be investigated by route and group. In comparing causal restraint under changing conditions, imputation or complete-case analysis should state assumptions and should not erase the likely direction of exclusion.[REF-09] [REF-19] [REF-38] [REF-52]

Causal interpretation of causal restraint under changing conditions should preserve baseline inequality, additional disruption, response differences and later change. For causal restraint under changing conditions, a post-disruption gap does not by itself identify its cause. In comparing causal restraint under changing conditions, evidence of exposure, timing and alternative explanations supports a bounded contribution claim; otherwise the report should describe association and the practical barrier requiring correction.[REF-19] [REF-38] [REF-52]

Public accountability for causal restraint under changing conditions should name the competent owner, resource, accessible correction and review date. In comparing causal restraint under changing conditions, learners need safe complaint and continuity routes. Within evidence on causal restraint under changing conditions, later review should ask whether the corrected service reached the group and whether participation, learning or wellbeing improved, while retaining the original definition and evidence limitations.[REF-09] [REF-19] [REF-38] [REF-52]

Accessible media and learning materials defines a particular comparison within the format accessibility. For accessible media and learning materials, the affected population is learners with visual, hearing, communication or cognitive access needs, and the immediate evidence concerns usable format and independent or supported access. In comparing accessible media and learning materials, a valid account states who was exposed, what institutional response was offered, who could use it and which learner condition was later observed. Within evidence on accessible media and learning materials, these stages should not collapse into a single indicator.[REF-10] [REF-11] [REF-15] [REF-49]

The principal risk is that standard materials are reported as available despite inaccessible format. In comparing accessible media and learning materials, that error can make excluded learners invisible or assign unequal effects to personal characteristics rather than institutions and material conditions. Within evidence on accessible media and learning materials, review should reconstruct entry into each source, actual service receipt and the barrier controlled by households, schools or public authorities.[REF-10] [REF-11] [REF-15]

The recommended response is to verify compatibility, communication and actual use. Within evidence on accessible media and learning materials, reporting should include counts, group levels, reference period, source and material missingness. For interpretation of accessible media and learning materials, a gap or ratio should appear beside absolute conditions so apparent parity does not conceal low provision for all. In decisions about accessible media and learning materials, where the denominator or category is unstable, the conclusion should remain bounded.[REF-11] [REF-15] [REF-49]

Evidence for accessible media and learning materials should distinguish administrative status, household access, actual contact, educational use and later outcome. For interpretation of accessible media and learning materials, surveys, registers, provider records and experience evidence observe different populations. In decisions about accessible media and learning materials, agreement can strengthen a finding, while disagreement may reveal excluded settings, timing or definition differences that require explanation.[REF-10] [REF-49]

Intersectional interpretation of accessible media and learning materials should select characteristics tied to a plausible barrier and public decision. In decisions about accessible media and learning materials, disability, gender, location and income can combine with age, language or migration. For correction of accessible media and learning materials, counts, intervals and disclosure risk should determine the precision of publication. For accessible media and learning materials, small groups require neither invisibility nor unwarranted certainty.[REF-11] [REF-15] [REF-49]

Missing evidence for accessible media and learning materials is itself potentially unequal. For correction of accessible media and learning materials, learners unable to access education media may also be unable to answer connected surveys or submit work. For accessible media and learning materials, non-response should therefore be investigated by route and group. In comparing accessible media and learning materials, imputation or complete-case analysis should state assumptions and should not erase the likely direction of exclusion.[REF-10] [REF-11] [REF-15] [REF-49]

Causal interpretation of accessible media and learning materials should preserve baseline inequality, additional disruption, response differences and later change. For accessible media and learning materials, a post-disruption gap does not by itself identify its cause. In comparing accessible media and learning materials, evidence of exposure, timing and alternative explanations supports a bounded contribution claim; otherwise the report should describe association and the practical barrier requiring correction.[REF-11] [REF-15] [REF-49]

Public accountability for accessible media and learning materials should name the competent owner, resource, accessible correction and review date. In comparing accessible media and learning materials, learners need safe complaint and continuity routes. Within evidence on accessible media and learning materials, later review should ask whether the corrected service reached the group and whether participation, learning or wellbeing improved, while retaining the original definition and evidence limitations.[REF-10] [REF-11] [REF-15] [REF-49]

Assistive resources and technical support defines a particular comparison within the assistive continuity. For assistive resources and technical support, the affected population is learners relying on devices, adapted materials or human assistance, and the immediate evidence concerns availability, maintenance and support. In comparing assistive resources and technical support, a valid account states who was exposed, what institutional response was offered, who could use it and which learner condition was later observed. Within evidence on assistive resources and technical support, these stages should not collapse into a single indicator.[REF-11] [REF-15] [REF-53] [REF-54]

The principal risk is that assistive resources remain at institutions or fail without support. In comparing assistive resources and technical support, that error can make excluded learners invisible or assign unequal effects to personal characteristics rather than institutions and material conditions. Within evidence on assistive resources and technical support, review should reconstruct entry into each source, actual service receipt and the barrier controlled by households, schools or public authorities.[REF-11] [REF-15] [REF-53]

The recommended response is to arrange safe access, maintenance and alternatives. Within evidence on assistive resources and technical support, reporting should include counts, group levels, reference period, source and material missingness. For interpretation of assistive resources and technical support, a gap or ratio should appear beside absolute conditions so apparent parity does not conceal low provision for all. In decisions about assistive resources and technical support, where the denominator or category is unstable, the conclusion should remain bounded.[REF-15] [REF-53] [REF-54]

Evidence for assistive resources and technical support should distinguish administrative status, household access, actual contact, educational use and later outcome. For interpretation of assistive resources and technical support, surveys, registers, provider records and experience evidence observe different populations. In decisions about assistive resources and technical support, agreement can strengthen a finding, while disagreement may reveal excluded settings, timing or definition differences that require explanation.[REF-11] [REF-54]

Intersectional interpretation of assistive resources and technical support should select characteristics tied to a plausible barrier and public decision. In decisions about assistive resources and technical support, disability, gender, location and income can combine with age, language or migration. For correction of assistive resources and technical support, counts, intervals and disclosure risk should determine the precision of publication. For assistive resources and technical support, small groups require neither invisibility nor unwarranted certainty.[REF-15] [REF-53] [REF-54]

Missing evidence for assistive resources and technical support is itself potentially unequal. For correction of assistive resources and technical support, learners unable to access education media may also be unable to answer connected surveys or submit work. For assistive resources and technical support, non-response should therefore be investigated by route and group. In comparing assistive resources and technical support, imputation or complete-case analysis should state assumptions and should not erase the likely direction of exclusion.[REF-11] [REF-15] [REF-53] [REF-54]

Causal interpretation of assistive resources and technical support should preserve baseline inequality, additional disruption, response differences and later change. For assistive resources and technical support, a post-disruption gap does not by itself identify its cause. In comparing assistive resources and technical support, evidence of exposure, timing and alternative explanations supports a bounded contribution claim; otherwise the report should describe association and the practical barrier requiring correction.[REF-15] [REF-53] [REF-54]

Public accountability for assistive resources and technical support should name the competent owner, resource, accessible correction and review date. In comparing assistive resources and technical support, learners need safe complaint and continuity routes. Within evidence on assistive resources and technical support, later review should ask whether the corrected service reached the group and whether participation, learning or wellbeing improved, while retaining the original definition and evidence limitations.[REF-11] [REF-15] [REF-53] [REF-54]

Specialist and therapeutic service interruption defines a particular comparison within the support interruption. For specialist and therapeutic service interruption, the affected population is learners receiving specialist education or linked services, and the immediate evidence concerns frequency, continuity and educational consequence. In comparing specialist and therapeutic service interruption, a valid account states who was exposed, what institutional response was offered, who could use it and which learner condition was later observed. Within evidence on specialist and therapeutic service interruption, these stages should not collapse into a single indicator.[REF-10] [REF-15] [REF-50] [REF-53]

The principal risk is that lost support is absent from remote-learning measures. In comparing specialist and therapeutic service interruption, that error can make excluded learners invisible or assign unequal effects to personal characteristics rather than institutions and material conditions. Within evidence on specialist and therapeutic service interruption, review should reconstruct entry into each source, actual service receipt and the barrier controlled by households, schools or public authorities.[REF-10] [REF-15] [REF-50]

The recommended response is to report interruption and substitute support received. Within evidence on specialist and therapeutic service interruption, reporting should include counts, group levels, reference period, source and material missingness. For interpretation of specialist and therapeutic service interruption, a gap or ratio should appear beside absolute conditions so apparent parity does not conceal low provision for all. In decisions about specialist and therapeutic service interruption, where the denominator or category is unstable, the conclusion should remain bounded.[REF-15] [REF-50] [REF-53]

Evidence for specialist and therapeutic service interruption should distinguish administrative status, household access, actual contact, educational use and later outcome. For interpretation of specialist and therapeutic service interruption, surveys, registers, provider records and experience evidence observe different populations. In decisions about specialist and therapeutic service interruption, agreement can strengthen a finding, while disagreement may reveal excluded settings, timing or definition differences that require explanation.[REF-10] [REF-53]

Intersectional interpretation of specialist and therapeutic service interruption should select characteristics tied to a plausible barrier and public decision. In decisions about specialist and therapeutic service interruption, disability, gender, location and income can combine with age, language or migration. For correction of specialist and therapeutic service interruption, counts, intervals and disclosure risk should determine the precision of publication. For specialist and therapeutic service interruption, small groups require neither invisibility nor unwarranted certainty.[REF-15] [REF-50] [REF-53]

Missing evidence for specialist and therapeutic service interruption is itself potentially unequal. For correction of specialist and therapeutic service interruption, learners unable to access education media may also be unable to answer connected surveys or submit work. For specialist and therapeutic service interruption, non-response should therefore be investigated by route and group. In comparing specialist and therapeutic service interruption, imputation or complete-case analysis should state assumptions and should not erase the likely direction of exclusion.[REF-10] [REF-15] [REF-50] [REF-53]

Causal interpretation of specialist and therapeutic service interruption should preserve baseline inequality, additional disruption, response differences and later change. For specialist and therapeutic service interruption, a post-disruption gap does not by itself identify its cause. In comparing specialist and therapeutic service interruption, evidence of exposure, timing and alternative explanations supports a bounded contribution claim; otherwise the report should describe association and the practical barrier requiring correction.[REF-15] [REF-50] [REF-53]

Public accountability for specialist and therapeutic service interruption should name the competent owner, resource, accessible correction and review date. In comparing specialist and therapeutic service interruption, learners need safe complaint and continuity routes. Within evidence on specialist and therapeutic service interruption, later review should ask whether the corrected service reached the group and whether participation, learning or wellbeing improved, while retaining the original definition and evidence limitations.[REF-10] [REF-15] [REF-50] [REF-53]

Caregiver burden and institutional duty defines a particular comparison within the support burden. For caregiver burden and institutional duty, the affected population is households asked to replace specialist or teaching roles, and the immediate evidence concerns adult time, skill and stress. In comparing caregiver burden and institutional duty, a valid account states who was exposed, what institutional response was offered, who could use it and which learner condition was later observed. Within evidence on caregiver burden and institutional duty, these stages should not collapse into a single indicator.[REF-12] [REF-49] [REF-52] [REF-55]

The principal risk is that remote education transfers professional responsibility to families. In comparing caregiver burden and institutional duty, that error can make excluded learners invisible or assign unequal effects to personal characteristics rather than institutions and material conditions. Within evidence on caregiver burden and institutional duty, review should reconstruct entry into each source, actual service receipt and the barrier controlled by households, schools or public authorities.[REF-12] [REF-49] [REF-52]

The recommended response is to provide reachable professional guidance and feasible tasks. Within evidence on caregiver burden and institutional duty, reporting should include counts, group levels, reference period, source and material missingness. For interpretation of caregiver burden and institutional duty, a gap or ratio should appear beside absolute conditions so apparent parity does not conceal low provision for all. In decisions about caregiver burden and institutional duty, where the denominator or category is unstable, the conclusion should remain bounded.[REF-49] [REF-52] [REF-55]

Evidence for caregiver burden and institutional duty should distinguish administrative status, household access, actual contact, educational use and later outcome. For interpretation of caregiver burden and institutional duty, surveys, registers, provider records and experience evidence observe different populations. In decisions about caregiver burden and institutional duty, agreement can strengthen a finding, while disagreement may reveal excluded settings, timing or definition differences that require explanation.[REF-12] [REF-55]

Intersectional interpretation of caregiver burden and institutional duty should select characteristics tied to a plausible barrier and public decision. In decisions about caregiver burden and institutional duty, disability, gender, location and income can combine with age, language or migration. For correction of caregiver burden and institutional duty, counts, intervals and disclosure risk should determine the precision of publication. For caregiver burden and institutional duty, small groups require neither invisibility nor unwarranted certainty.[REF-49] [REF-52] [REF-55]

Missing evidence for caregiver burden and institutional duty is itself potentially unequal. For correction of caregiver burden and institutional duty, learners unable to access education media may also be unable to answer connected surveys or submit work. For caregiver burden and institutional duty, non-response should therefore be investigated by route and group. In comparing caregiver burden and institutional duty, imputation or complete-case analysis should state assumptions and should not erase the likely direction of exclusion.[REF-12] [REF-49] [REF-52] [REF-55]

Causal interpretation of caregiver burden and institutional duty should preserve baseline inequality, additional disruption, response differences and later change. For caregiver burden and institutional duty, a post-disruption gap does not by itself identify its cause. In comparing caregiver burden and institutional duty, evidence of exposure, timing and alternative explanations supports a bounded contribution claim; otherwise the report should describe association and the practical barrier requiring correction.[REF-49] [REF-52] [REF-55]

Public accountability for caregiver burden and institutional duty should name the competent owner, resource, accessible correction and review date. In comparing caregiver burden and institutional duty, learners need safe complaint and continuity routes. Within evidence on caregiver burden and institutional duty, later review should ask whether the corrected service reached the group and whether participation, learning or wellbeing improved, while retaining the original definition and evidence limitations.[REF-12] [REF-49] [REF-52] [REF-55]

Accessible assessment and participation evidence defines a particular comparison within the measurement accessibility. For accessible assessment and participation evidence, the affected population is learners included or excluded from remote checks, and the immediate evidence concerns accommodation, response route and missingness. In comparing accessible assessment and participation evidence, a valid account states who was exposed, what institutional response was offered, who could use it and which learner condition was later observed. Within evidence on accessible assessment and participation evidence, these stages should not collapse into a single indicator.[REF-09] [REF-15] [REF-17] [REF-38]

The principal risk is that inaccessible evidence collection makes learners statistically invisible. In comparing accessible assessment and participation evidence, that error can make excluded learners invisible or assign unequal effects to personal characteristics rather than institutions and material conditions. Within evidence on accessible assessment and participation evidence, review should reconstruct entry into each source, actual service receipt and the barrier controlled by households, schools or public authorities.[REF-09] [REF-15] [REF-17]

The recommended response is to offer accommodations and report exclusions. Within evidence on accessible assessment and participation evidence, reporting should include counts, group levels, reference period, source and material missingness. For interpretation of accessible assessment and participation evidence, a gap or ratio should appear beside absolute conditions so apparent parity does not conceal low provision for all. In decisions about accessible assessment and participation evidence, where the denominator or category is unstable, the conclusion should remain bounded.[REF-15] [REF-17] [REF-38]

Evidence for accessible assessment and participation evidence should distinguish administrative status, household access, actual contact, educational use and later outcome. For interpretation of accessible assessment and participation evidence, surveys, registers, provider records and experience evidence observe different populations. In decisions about accessible assessment and participation evidence, agreement can strengthen a finding, while disagreement may reveal excluded settings, timing or definition differences that require explanation.[REF-09] [REF-38]

Intersectional interpretation of accessible assessment and participation evidence should select characteristics tied to a plausible barrier and public decision. In decisions about accessible assessment and participation evidence, disability, gender, location and income can combine with age, language or migration. For correction of accessible assessment and participation evidence, counts, intervals and disclosure risk should determine the precision of publication. For accessible assessment and participation evidence, small groups require neither invisibility nor unwarranted certainty.[REF-15] [REF-17] [REF-38]

Missing evidence for accessible assessment and participation evidence is itself potentially unequal. For correction of accessible assessment and participation evidence, learners unable to access education media may also be unable to answer connected surveys or submit work. For accessible assessment and participation evidence, non-response should therefore be investigated by route and group. In comparing accessible assessment and participation evidence, imputation or complete-case analysis should state assumptions and should not erase the likely direction of exclusion.[REF-09] [REF-15] [REF-17] [REF-38]

Causal interpretation of accessible assessment and participation evidence should preserve baseline inequality, additional disruption, response differences and later change. For accessible assessment and participation evidence, a post-disruption gap does not by itself identify its cause. In comparing accessible assessment and participation evidence, evidence of exposure, timing and alternative explanations supports a bounded contribution claim; otherwise the report should describe association and the practical barrier requiring correction.[REF-15] [REF-17] [REF-38]

Public accountability for accessible assessment and participation evidence should name the competent owner, resource, accessible correction and review date. In comparing accessible assessment and participation evidence, learners need safe complaint and continuity routes. Within evidence on accessible assessment and participation evidence, later review should ask whether the corrected service reached the group and whether participation, learning or wellbeing improved, while retaining the original definition and evidence limitations.[REF-09] [REF-15] [REF-17] [REF-38]

Return, accommodation and continuity defines a particular comparison within the re-entry inclusion. For return, accommodation and continuity, the affected population is learners returning with changed support needs, and the immediate evidence concerns safe access, accommodation and restored service. In comparing return, accommodation and continuity, a valid account states who was exposed, what institutional response was offered, who could use it and which learner condition was later observed. Within evidence on return, accommodation and continuity, these stages should not collapse into a single indicator.[REF-11] [REF-15] [REF-50] [REF-53]

The principal risk is that reopening is counted although necessary support remains unavailable. In comparing return, accommodation and continuity, that error can make excluded learners invisible or assign unequal effects to personal characteristics rather than institutions and material conditions. Within evidence on return, accommodation and continuity, review should reconstruct entry into each source, actual service receipt and the barrier controlled by households, schools or public authorities.[REF-11] [REF-15] [REF-50]

The recommended response is to verify accessibility and service restoration. Within evidence on return, accommodation and continuity, reporting should include counts, group levels, reference period, source and material missingness. For interpretation of return, accommodation and continuity, a gap or ratio should appear beside absolute conditions so apparent parity does not conceal low provision for all. In decisions about return, accommodation and continuity, where the denominator or category is unstable, the conclusion should remain bounded.[REF-15] [REF-50] [REF-53]

Evidence for return, accommodation and continuity should distinguish administrative status, household access, actual contact, educational use and later outcome. For interpretation of return, accommodation and continuity, surveys, registers, provider records and experience evidence observe different populations. In decisions about return, accommodation and continuity, agreement can strengthen a finding, while disagreement may reveal excluded settings, timing or definition differences that require explanation.[REF-11] [REF-53]

Intersectional interpretation of return, accommodation and continuity should select characteristics tied to a plausible barrier and public decision. In decisions about return, accommodation and continuity, disability, gender, location and income can combine with age, language or migration. For correction of return, accommodation and continuity, counts, intervals and disclosure risk should determine the precision of publication. For return, accommodation and continuity, small groups require neither invisibility nor unwarranted certainty.[REF-15] [REF-50] [REF-53]

Missing evidence for return, accommodation and continuity is itself potentially unequal. For correction of return, accommodation and continuity, learners unable to access education media may also be unable to answer connected surveys or submit work. For return, accommodation and continuity, non-response should therefore be investigated by route and group. In comparing return, accommodation and continuity, imputation or complete-case analysis should state assumptions and should not erase the likely direction of exclusion.[REF-11] [REF-15] [REF-50] [REF-53]

Causal interpretation of return, accommodation and continuity should preserve baseline inequality, additional disruption, response differences and later change. For return, accommodation and continuity, a post-disruption gap does not by itself identify its cause. In comparing return, accommodation and continuity, evidence of exposure, timing and alternative explanations supports a bounded contribution claim; otherwise the report should describe association and the practical barrier requiring correction.[REF-15] [REF-50] [REF-53]

Public accountability for return, accommodation and continuity should name the competent owner, resource, accessible correction and review date. In comparing return, accommodation and continuity, learners need safe complaint and continuity routes. Within evidence on return, accommodation and continuity, later review should ask whether the corrected service reached the group and whether participation, learning or wellbeing improved, while retaining the original definition and evidence limitations.[REF-11] [REF-15] [REF-50] [REF-53]

Device priority and household allocation defines a particular comparison within the gendered access. For device priority and household allocation, the affected population is learners sharing scarce devices and study time, and the immediate evidence concerns who controls access, when and for how long. In comparing device priority and household allocation, a valid account states who was exposed, what institutional response was offered, who could use it and which learner condition was later observed. Within evidence on device priority and household allocation, these stages should not collapse into a single indicator.[REF-07] [REF-38] [REF-55] [REF-56]

The principal risk is that household device presence is assumed to benefit learners equally. In comparing device priority and household allocation, that error can make excluded learners invisible or assign unequal effects to personal characteristics rather than institutions and material conditions. Within evidence on device priority and household allocation, review should reconstruct entry into each source, actual service receipt and the barrier controlled by households, schools or public authorities.[REF-07] [REF-38] [REF-55]

The recommended response is to measure individual usability and allocation. Within evidence on device priority and household allocation, reporting should include counts, group levels, reference period, source and material missingness. For interpretation of device priority and household allocation, a gap or ratio should appear beside absolute conditions so apparent parity does not conceal low provision for all. In decisions about device priority and household allocation, where the denominator or category is unstable, the conclusion should remain bounded.[REF-38] [REF-55] [REF-56]

Evidence for device priority and household allocation should distinguish administrative status, household access, actual contact, educational use and later outcome. For interpretation of device priority and household allocation, surveys, registers, provider records and experience evidence observe different populations. In decisions about device priority and household allocation, agreement can strengthen a finding, while disagreement may reveal excluded settings, timing or definition differences that require explanation.[REF-07] [REF-56]

Intersectional interpretation of device priority and household allocation should select characteristics tied to a plausible barrier and public decision. In decisions about device priority and household allocation, disability, gender, location and income can combine with age, language or migration. For correction of device priority and household allocation, counts, intervals and disclosure risk should determine the precision of publication. For device priority and household allocation, small groups require neither invisibility nor unwarranted certainty.[REF-38] [REF-55] [REF-56]

Missing evidence for device priority and household allocation is itself potentially unequal. For correction of device priority and household allocation, learners unable to access education media may also be unable to answer connected surveys or submit work. For device priority and household allocation, non-response should therefore be investigated by route and group. In comparing device priority and household allocation, imputation or complete-case analysis should state assumptions and should not erase the likely direction of exclusion.[REF-07] [REF-38] [REF-55] [REF-56]

Causal interpretation of device priority and household allocation should preserve baseline inequality, additional disruption, response differences and later change. For device priority and household allocation, a post-disruption gap does not by itself identify its cause. In comparing device priority and household allocation, evidence of exposure, timing and alternative explanations supports a bounded contribution claim; otherwise the report should describe association and the practical barrier requiring correction.[REF-38] [REF-55] [REF-56]

Public accountability for device priority and household allocation should name the competent owner, resource, accessible correction and review date. In comparing device priority and household allocation, learners need safe complaint and continuity routes. Within evidence on device priority and household allocation, later review should ask whether the corrected service reached the group and whether participation, learning or wellbeing improved, while retaining the original definition and evidence limitations.[REF-07] [REF-38] [REF-55] [REF-56]

Care, domestic work and study time defines a particular comparison within the time burden. For care, domestic work and study time, the affected population is learners carrying different household responsibilities, and the immediate evidence concerns available time, interruption and fatigue. In comparing care, domestic work and study time, a valid account states who was exposed, what institutional response was offered, who could use it and which learner condition was later observed. Within evidence on care, domestic work and study time, these stages should not collapse into a single indicator.[REF-03] [REF-12] [REF-52] [REF-53]

The principal risk is that task non-completion is interpreted without unpaid care evidence. In comparing care, domestic work and study time, that error can make excluded learners invisible or assign unequal effects to personal characteristics rather than institutions and material conditions. Within evidence on care, domestic work and study time, review should reconstruct entry into each source, actual service receipt and the barrier controlled by households, schools or public authorities.[REF-03] [REF-12] [REF-52]

The recommended response is to report time constraints and design flexible support. Within evidence on care, domestic work and study time, reporting should include counts, group levels, reference period, source and material missingness. For interpretation of care, domestic work and study time, a gap or ratio should appear beside absolute conditions so apparent parity does not conceal low provision for all. In decisions about care, domestic work and study time, where the denominator or category is unstable, the conclusion should remain bounded.[REF-12] [REF-52] [REF-53]

Evidence for care, domestic work and study time should distinguish administrative status, household access, actual contact, educational use and later outcome. For interpretation of care, domestic work and study time, surveys, registers, provider records and experience evidence observe different populations. In decisions about care, domestic work and study time, agreement can strengthen a finding, while disagreement may reveal excluded settings, timing or definition differences that require explanation.[REF-03] [REF-53]

Intersectional interpretation of care, domestic work and study time should select characteristics tied to a plausible barrier and public decision. In decisions about care, domestic work and study time, disability, gender, location and income can combine with age, language or migration. For correction of care, domestic work and study time, counts, intervals and disclosure risk should determine the precision of publication. For care, domestic work and study time, small groups require neither invisibility nor unwarranted certainty.[REF-12] [REF-52] [REF-53]

Missing evidence for care, domestic work and study time is itself potentially unequal. For correction of care, domestic work and study time, learners unable to access education media may also be unable to answer connected surveys or submit work. For care, domestic work and study time, non-response should therefore be investigated by route and group. In comparing care, domestic work and study time, imputation or complete-case analysis should state assumptions and should not erase the likely direction of exclusion.[REF-03] [REF-12] [REF-52] [REF-53]

Causal interpretation of care, domestic work and study time should preserve baseline inequality, additional disruption, response differences and later change. For care, domestic work and study time, a post-disruption gap does not by itself identify its cause. In comparing care, domestic work and study time, evidence of exposure, timing and alternative explanations supports a bounded contribution claim; otherwise the report should describe association and the practical barrier requiring correction.[REF-12] [REF-52] [REF-53]

Public accountability for care, domestic work and study time should name the competent owner, resource, accessible correction and review date. In comparing care, domestic work and study time, learners need safe complaint and continuity routes. Within evidence on care, domestic work and study time, later review should ask whether the corrected service reached the group and whether participation, learning or wellbeing improved, while retaining the original definition and evidence limitations.[REF-03] [REF-12] [REF-52] [REF-53]

Safety, harassment and harmful contact defines a particular comparison within the protection risk. For safety, harassment and harmful contact, the affected population is learners exposed to violence or unsafe communication, and the immediate evidence concerns experience, report route and response. In comparing safety, harassment and harmful contact, a valid account states who was exposed, what institutional response was offered, who could use it and which learner condition was later observed. Within evidence on safety, harassment and harmful contact, these stages should not collapse into a single indicator.[REF-14] [REF-20] [REF-21] [REF-50]

The principal risk is that absence of reports is treated as safety. In comparing safety, harassment and harmful contact, that error can make excluded learners invisible or assign unequal effects to personal characteristics rather than institutions and material conditions. Within evidence on safety, harassment and harmful contact, review should reconstruct entry into each source, actual service receipt and the barrier controlled by households, schools or public authorities.[REF-14] [REF-20] [REF-21]

The recommended response is to provide confidential contact and non-retaliation. Within evidence on safety, harassment and harmful contact, reporting should include counts, group levels, reference period, source and material missingness. For interpretation of safety, harassment and harmful contact, a gap or ratio should appear beside absolute conditions so apparent parity does not conceal low provision for all. In decisions about safety, harassment and harmful contact, where the denominator or category is unstable, the conclusion should remain bounded.[REF-20] [REF-21] [REF-50]

Evidence for safety, harassment and harmful contact should distinguish administrative status, household access, actual contact, educational use and later outcome. For interpretation of safety, harassment and harmful contact, surveys, registers, provider records and experience evidence observe different populations. In decisions about safety, harassment and harmful contact, agreement can strengthen a finding, while disagreement may reveal excluded settings, timing or definition differences that require explanation.[REF-14] [REF-50]

Intersectional interpretation of safety, harassment and harmful contact should select characteristics tied to a plausible barrier and public decision. In decisions about safety, harassment and harmful contact, disability, gender, location and income can combine with age, language or migration. For correction of safety, harassment and harmful contact, counts, intervals and disclosure risk should determine the precision of publication. For safety, harassment and harmful contact, small groups require neither invisibility nor unwarranted certainty.[REF-20] [REF-21] [REF-50]

Missing evidence for safety, harassment and harmful contact is itself potentially unequal. For correction of safety, harassment and harmful contact, learners unable to access education media may also be unable to answer connected surveys or submit work. For safety, harassment and harmful contact, non-response should therefore be investigated by route and group. In comparing safety, harassment and harmful contact, imputation or complete-case analysis should state assumptions and should not erase the likely direction of exclusion.[REF-14] [REF-20] [REF-21] [REF-50]

Causal interpretation of safety, harassment and harmful contact should preserve baseline inequality, additional disruption, response differences and later change. For safety, harassment and harmful contact, a post-disruption gap does not by itself identify its cause. In comparing safety, harassment and harmful contact, evidence of exposure, timing and alternative explanations supports a bounded contribution claim; otherwise the report should describe association and the practical barrier requiring correction.[REF-20] [REF-21] [REF-50]

Public accountability for safety, harassment and harmful contact should name the competent owner, resource, accessible correction and review date. In comparing safety, harassment and harmful contact, learners need safe complaint and continuity routes. Within evidence on safety, harassment and harmful contact, later review should ask whether the corrected service reached the group and whether participation, learning or wellbeing improved, while retaining the original definition and evidence limitations.[REF-14] [REF-20] [REF-21] [REF-50]

Economic pressure and early leaving defines a particular comparison within the continuity pressure. For economic pressure and early leaving, the affected population is adolescents entering work, care or early union under household stress, and the immediate evidence concerns continued contact, work and return. In comparing economic pressure and early leaving, a valid account states who was exposed, what institutional response was offered, who could use it and which learner condition was later observed. Within evidence on economic pressure and early leaving, these stages should not collapse into a single indicator.[REF-03] [REF-23] [REF-35] [REF-52]

The principal risk is that temporary absence becomes durable leaving without follow-up. In comparing economic pressure and early leaving, that error can make excluded learners invisible or assign unequal effects to personal characteristics rather than institutions and material conditions. Within evidence on economic pressure and early leaving, review should reconstruct entry into each source, actual service receipt and the barrier controlled by households, schools or public authorities.[REF-03] [REF-23] [REF-35]

The recommended response is to maintain contact and finance re-entry support. Within evidence on economic pressure and early leaving, reporting should include counts, group levels, reference period, source and material missingness. For interpretation of economic pressure and early leaving, a gap or ratio should appear beside absolute conditions so apparent parity does not conceal low provision for all. In decisions about economic pressure and early leaving, where the denominator or category is unstable, the conclusion should remain bounded.[REF-23] [REF-35] [REF-52]

Evidence for economic pressure and early leaving should distinguish administrative status, household access, actual contact, educational use and later outcome. For interpretation of economic pressure and early leaving, surveys, registers, provider records and experience evidence observe different populations. In decisions about economic pressure and early leaving, agreement can strengthen a finding, while disagreement may reveal excluded settings, timing or definition differences that require explanation.[REF-03] [REF-52]

Intersectional interpretation of economic pressure and early leaving should select characteristics tied to a plausible barrier and public decision. In decisions about economic pressure and early leaving, disability, gender, location and income can combine with age, language or migration. For correction of economic pressure and early leaving, counts, intervals and disclosure risk should determine the precision of publication. For economic pressure and early leaving, small groups require neither invisibility nor unwarranted certainty.[REF-23] [REF-35] [REF-52]

Missing evidence for economic pressure and early leaving is itself potentially unequal. For correction of economic pressure and early leaving, learners unable to access education media may also be unable to answer connected surveys or submit work. For economic pressure and early leaving, non-response should therefore be investigated by route and group. In comparing economic pressure and early leaving, imputation or complete-case analysis should state assumptions and should not erase the likely direction of exclusion.[REF-03] [REF-23] [REF-35] [REF-52]

Causal interpretation of economic pressure and early leaving should preserve baseline inequality, additional disruption, response differences and later change. For economic pressure and early leaving, a post-disruption gap does not by itself identify its cause. In comparing economic pressure and early leaving, evidence of exposure, timing and alternative explanations supports a bounded contribution claim; otherwise the report should describe association and the practical barrier requiring correction.[REF-23] [REF-35] [REF-52]

Public accountability for economic pressure and early leaving should name the competent owner, resource, accessible correction and review date. In comparing economic pressure and early leaving, learners need safe complaint and continuity routes. Within evidence on economic pressure and early leaving, later review should ask whether the corrected service reached the group and whether participation, learning or wellbeing improved, while retaining the original definition and evidence limitations.[REF-03] [REF-23] [REF-35] [REF-52]

Pregnancy, parenthood and re-entry defines a particular comparison within the return barrier. For pregnancy, parenthood and re-entry, the affected population is pregnant learners and young parents, and the immediate evidence concerns continued education and return conditions. In comparing pregnancy, parenthood and re-entry, a valid account states who was exposed, what institutional response was offered, who could use it and which learner condition was later observed. Within evidence on pregnancy, parenthood and re-entry, these stages should not collapse into a single indicator.[REF-10] [REF-14] [REF-22] [REF-53]

The principal risk is that disruption compounds exclusionary rules or practical barriers. In comparing pregnancy, parenthood and re-entry, that error can make excluded learners invisible or assign unequal effects to personal characteristics rather than institutions and material conditions. Within evidence on pregnancy, parenthood and re-entry, review should reconstruct entry into each source, actual service receipt and the barrier controlled by households, schools or public authorities.[REF-10] [REF-14] [REF-22]

The recommended response is to protect non-discriminatory continuation and flexible return. Within evidence on pregnancy, parenthood and re-entry, reporting should include counts, group levels, reference period, source and material missingness. For interpretation of pregnancy, parenthood and re-entry, a gap or ratio should appear beside absolute conditions so apparent parity does not conceal low provision for all. In decisions about pregnancy, parenthood and re-entry, where the denominator or category is unstable, the conclusion should remain bounded.[REF-14] [REF-22] [REF-53]

Evidence for pregnancy, parenthood and re-entry should distinguish administrative status, household access, actual contact, educational use and later outcome. For interpretation of pregnancy, parenthood and re-entry, surveys, registers, provider records and experience evidence observe different populations. In decisions about pregnancy, parenthood and re-entry, agreement can strengthen a finding, while disagreement may reveal excluded settings, timing or definition differences that require explanation.[REF-10] [REF-53]

Intersectional interpretation of pregnancy, parenthood and re-entry should select characteristics tied to a plausible barrier and public decision. In decisions about pregnancy, parenthood and re-entry, disability, gender, location and income can combine with age, language or migration. For correction of pregnancy, parenthood and re-entry, counts, intervals and disclosure risk should determine the precision of publication. For pregnancy, parenthood and re-entry, small groups require neither invisibility nor unwarranted certainty.[REF-14] [REF-22] [REF-53]

Missing evidence for pregnancy, parenthood and re-entry is itself potentially unequal. For correction of pregnancy, parenthood and re-entry, learners unable to access education media may also be unable to answer connected surveys or submit work. For pregnancy, parenthood and re-entry, non-response should therefore be investigated by route and group. In comparing pregnancy, parenthood and re-entry, imputation or complete-case analysis should state assumptions and should not erase the likely direction of exclusion.[REF-10] [REF-14] [REF-22] [REF-53]

Causal interpretation of pregnancy, parenthood and re-entry should preserve baseline inequality, additional disruption, response differences and later change. For pregnancy, parenthood and re-entry, a post-disruption gap does not by itself identify its cause. In comparing pregnancy, parenthood and re-entry, evidence of exposure, timing and alternative explanations supports a bounded contribution claim; otherwise the report should describe association and the practical barrier requiring correction.[REF-14] [REF-22] [REF-53]

Public accountability for pregnancy, parenthood and re-entry should name the competent owner, resource, accessible correction and review date. In comparing pregnancy, parenthood and re-entry, learners need safe complaint and continuity routes. Within evidence on pregnancy, parenthood and re-entry, later review should ask whether the corrected service reached the group and whether participation, learning or wellbeing improved, while retaining the original definition and evidence limitations.[REF-10] [REF-14] [REF-22] [REF-53]

Beyond binary averages defines a particular comparison within the gender interpretation. For beyond binary averages, the affected population is learners whose experience is not captured by broad categories, and the immediate evidence concerns group levels, local context and safety. In comparing beyond binary averages, a valid account states who was exposed, what institutional response was offered, who could use it and which learner condition was later observed. Within evidence on beyond binary averages, these stages should not collapse into a single indicator.[REF-09] [REF-38] [REF-44] [REF-53]

The principal risk is that national parity conceals universally weak provision or local opposition. In comparing beyond binary averages, that error can make excluded learners invisible or assign unequal effects to personal characteristics rather than institutions and material conditions. Within evidence on beyond binary averages, review should reconstruct entry into each source, actual service receipt and the barrier controlled by households, schools or public authorities.[REF-09] [REF-38] [REF-44]

The recommended response is to publish levels and contextual mechanisms. Within evidence on beyond binary averages, reporting should include counts, group levels, reference period, source and material missingness. For interpretation of beyond binary averages, a gap or ratio should appear beside absolute conditions so apparent parity does not conceal low provision for all. In decisions about beyond binary averages, where the denominator or category is unstable, the conclusion should remain bounded.[REF-38] [REF-44] [REF-53]

Evidence for beyond binary averages should distinguish administrative status, household access, actual contact, educational use and later outcome. For interpretation of beyond binary averages, surveys, registers, provider records and experience evidence observe different populations. In decisions about beyond binary averages, agreement can strengthen a finding, while disagreement may reveal excluded settings, timing or definition differences that require explanation.[REF-09] [REF-53]

Intersectional interpretation of beyond binary averages should select characteristics tied to a plausible barrier and public decision. In decisions about beyond binary averages, disability, gender, location and income can combine with age, language or migration. For correction of beyond binary averages, counts, intervals and disclosure risk should determine the precision of publication. For beyond binary averages, small groups require neither invisibility nor unwarranted certainty.[REF-38] [REF-44] [REF-53]

Missing evidence for beyond binary averages is itself potentially unequal. For correction of beyond binary averages, learners unable to access education media may also be unable to answer connected surveys or submit work. For beyond binary averages, non-response should therefore be investigated by route and group. In comparing beyond binary averages, imputation or complete-case analysis should state assumptions and should not erase the likely direction of exclusion.[REF-09] [REF-38] [REF-44] [REF-53]

Causal interpretation of beyond binary averages should preserve baseline inequality, additional disruption, response differences and later change. For beyond binary averages, a post-disruption gap does not by itself identify its cause. In comparing beyond binary averages, evidence of exposure, timing and alternative explanations supports a bounded contribution claim; otherwise the report should describe association and the practical barrier requiring correction.[REF-38] [REF-44] [REF-53]

Public accountability for beyond binary averages should name the competent owner, resource, accessible correction and review date. In comparing beyond binary averages, learners need safe complaint and continuity routes. Within evidence on beyond binary averages, later review should ask whether the corrected service reached the group and whether participation, learning or wellbeing improved, while retaining the original definition and evidence limitations.[REF-09] [REF-38] [REF-44] [REF-53]

Electricity, connectivity and device geography defines a particular comparison within the infrastructure access. For electricity, connectivity and device geography, the affected population is learners in places with varied infrastructure, and the immediate evidence concerns reliable service, device and recurring cost. In comparing electricity, connectivity and device geography, a valid account states who was exposed, what institutional response was offered, who could use it and which learner condition was later observed. Within evidence on electricity, connectivity and device geography, these stages should not collapse into a single indicator.[REF-07] [REF-11] [REF-49] [REF-55]

The principal risk is that nominal coverage is treated as continuous usable access. In comparing electricity, connectivity and device geography, that error can make excluded learners invisible or assign unequal effects to personal characteristics rather than institutions and material conditions. Within evidence on electricity, connectivity and device geography, review should reconstruct entry into each source, actual service receipt and the barrier controlled by households, schools or public authorities.[REF-07] [REF-11] [REF-49]

The recommended response is to report reliability, affordability and household availability. Within evidence on electricity, connectivity and device geography, reporting should include counts, group levels, reference period, source and material missingness. For interpretation of electricity, connectivity and device geography, a gap or ratio should appear beside absolute conditions so apparent parity does not conceal low provision for all. In decisions about electricity, connectivity and device geography, where the denominator or category is unstable, the conclusion should remain bounded.[REF-11] [REF-49] [REF-55]

Evidence for electricity, connectivity and device geography should distinguish administrative status, household access, actual contact, educational use and later outcome. For interpretation of electricity, connectivity and device geography, surveys, registers, provider records and experience evidence observe different populations. In decisions about electricity, connectivity and device geography, agreement can strengthen a finding, while disagreement may reveal excluded settings, timing or definition differences that require explanation.[REF-07] [REF-55]

Intersectional interpretation of electricity, connectivity and device geography should select characteristics tied to a plausible barrier and public decision. In decisions about electricity, connectivity and device geography, disability, gender, location and income can combine with age, language or migration. For correction of electricity, connectivity and device geography, counts, intervals and disclosure risk should determine the precision of publication. For electricity, connectivity and device geography, small groups require neither invisibility nor unwarranted certainty.[REF-11] [REF-49] [REF-55]

Missing evidence for electricity, connectivity and device geography is itself potentially unequal. For correction of electricity, connectivity and device geography, learners unable to access education media may also be unable to answer connected surveys or submit work. For electricity, connectivity and device geography, non-response should therefore be investigated by route and group. In comparing electricity, connectivity and device geography, imputation or complete-case analysis should state assumptions and should not erase the likely direction of exclusion.[REF-07] [REF-11] [REF-49] [REF-55]

Causal interpretation of electricity, connectivity and device geography should preserve baseline inequality, additional disruption, response differences and later change. For electricity, connectivity and device geography, a post-disruption gap does not by itself identify its cause. In comparing electricity, connectivity and device geography, evidence of exposure, timing and alternative explanations supports a bounded contribution claim; otherwise the report should describe association and the practical barrier requiring correction.[REF-11] [REF-49] [REF-55]

Public accountability for electricity, connectivity and device geography should name the competent owner, resource, accessible correction and review date. In comparing electricity, connectivity and device geography, learners need safe complaint and continuity routes. Within evidence on electricity, connectivity and device geography, later review should ask whether the corrected service reached the group and whether participation, learning or wellbeing improved, while retaining the original definition and evidence limitations.[REF-07] [REF-11] [REF-49] [REF-55]

Print, radio and television reach defines a particular comparison within the alternative-media reach. For print, radio and television reach, the affected population is learners using broadcast or physical materials, and the immediate evidence concerns signal, schedule, collection and feedback. In comparing print, radio and television reach, a valid account states who was exposed, what institutional response was offered, who could use it and which learner condition was later observed. Within evidence on print, radio and television reach, these stages should not collapse into a single indicator.[REF-05] [REF-48] [REF-49] [REF-56]

The principal risk is that non-connected alternatives are assumed to reach everyone excluded online. In comparing print, radio and television reach, that error can make excluded learners invisible or assign unequal effects to personal characteristics rather than institutions and material conditions. Within evidence on print, radio and television reach, review should reconstruct entry into each source, actual service receipt and the barrier controlled by households, schools or public authorities.[REF-05] [REF-48] [REF-49]

The recommended response is to verify receipt and teacher linkage. Within evidence on print, radio and television reach, reporting should include counts, group levels, reference period, source and material missingness. For interpretation of print, radio and television reach, a gap or ratio should appear beside absolute conditions so apparent parity does not conceal low provision for all. In decisions about print, radio and television reach, where the denominator or category is unstable, the conclusion should remain bounded.[REF-48] [REF-49] [REF-56]

Evidence for print, radio and television reach should distinguish administrative status, household access, actual contact, educational use and later outcome. For interpretation of print, radio and television reach, surveys, registers, provider records and experience evidence observe different populations. In decisions about print, radio and television reach, agreement can strengthen a finding, while disagreement may reveal excluded settings, timing or definition differences that require explanation.[REF-05] [REF-56]

Intersectional interpretation of print, radio and television reach should select characteristics tied to a plausible barrier and public decision. In decisions about print, radio and television reach, disability, gender, location and income can combine with age, language or migration. For correction of print, radio and television reach, counts, intervals and disclosure risk should determine the precision of publication. For print, radio and television reach, small groups require neither invisibility nor unwarranted certainty.[REF-48] [REF-49] [REF-56]

Missing evidence for print, radio and television reach is itself potentially unequal. For correction of print, radio and television reach, learners unable to access education media may also be unable to answer connected surveys or submit work. For print, radio and television reach, non-response should therefore be investigated by route and group. In comparing print, radio and television reach, imputation or complete-case analysis should state assumptions and should not erase the likely direction of exclusion.[REF-05] [REF-48] [REF-49] [REF-56]

Causal interpretation of print, radio and television reach should preserve baseline inequality, additional disruption, response differences and later change. For print, radio and television reach, a post-disruption gap does not by itself identify its cause. In comparing print, radio and television reach, evidence of exposure, timing and alternative explanations supports a bounded contribution claim; otherwise the report should describe association and the practical barrier requiring correction.[REF-48] [REF-49] [REF-56]

Public accountability for print, radio and television reach should name the competent owner, resource, accessible correction and review date. In comparing print, radio and television reach, learners need safe complaint and continuity routes. Within evidence on print, radio and television reach, later review should ask whether the corrected service reached the group and whether participation, learning or wellbeing improved, while retaining the original definition and evidence limitations.[REF-05] [REF-48] [REF-49] [REF-56]

Remote and small-school staffing defines a particular comparison within the territorial capacity. For remote and small-school staffing, the affected population is learners served by small or isolated institutions, and the immediate evidence concerns teacher availability, contact and substitute capacity. In comparing remote and small-school staffing, a valid account states who was exposed, what institutional response was offered, who could use it and which learner condition was later observed. Within evidence on remote and small-school staffing, these stages should not collapse into a single indicator.[REF-18] [REF-28] [REF-36] [REF-53]

The principal risk is that national remote plans ignore local staffing fragility. In comparing remote and small-school staffing, that error can make excluded learners invisible or assign unequal effects to personal characteristics rather than institutions and material conditions. Within evidence on remote and small-school staffing, review should reconstruct entry into each source, actual service receipt and the barrier controlled by households, schools or public authorities.[REF-18] [REF-28] [REF-36]

The recommended response is to report staff continuity and resource local solutions. Within evidence on remote and small-school staffing, reporting should include counts, group levels, reference period, source and material missingness. For interpretation of remote and small-school staffing, a gap or ratio should appear beside absolute conditions so apparent parity does not conceal low provision for all. In decisions about remote and small-school staffing, where the denominator or category is unstable, the conclusion should remain bounded.[REF-28] [REF-36] [REF-53]

Evidence for remote and small-school staffing should distinguish administrative status, household access, actual contact, educational use and later outcome. For interpretation of remote and small-school staffing, surveys, registers, provider records and experience evidence observe different populations. In decisions about remote and small-school staffing, agreement can strengthen a finding, while disagreement may reveal excluded settings, timing or definition differences that require explanation.[REF-18] [REF-53]

Intersectional interpretation of remote and small-school staffing should select characteristics tied to a plausible barrier and public decision. In decisions about remote and small-school staffing, disability, gender, location and income can combine with age, language or migration. For correction of remote and small-school staffing, counts, intervals and disclosure risk should determine the precision of publication. For remote and small-school staffing, small groups require neither invisibility nor unwarranted certainty.[REF-28] [REF-36] [REF-53]

Missing evidence for remote and small-school staffing is itself potentially unequal. For correction of remote and small-school staffing, learners unable to access education media may also be unable to answer connected surveys or submit work. For remote and small-school staffing, non-response should therefore be investigated by route and group. In comparing remote and small-school staffing, imputation or complete-case analysis should state assumptions and should not erase the likely direction of exclusion.[REF-18] [REF-28] [REF-36] [REF-53]

Causal interpretation of remote and small-school staffing should preserve baseline inequality, additional disruption, response differences and later change. For remote and small-school staffing, a post-disruption gap does not by itself identify its cause. In comparing remote and small-school staffing, evidence of exposure, timing and alternative explanations supports a bounded contribution claim; otherwise the report should describe association and the practical barrier requiring correction.[REF-28] [REF-36] [REF-53]

Public accountability for remote and small-school staffing should name the competent owner, resource, accessible correction and review date. In comparing remote and small-school staffing, learners need safe complaint and continuity routes. Within evidence on remote and small-school staffing, later review should ask whether the corrected service reached the group and whether participation, learning or wellbeing improved, while retaining the original definition and evidence limitations.[REF-18] [REF-28] [REF-36] [REF-53]

Informal settlements and unstable housing defines a particular comparison within the settlement condition. For informal settlements and unstable housing, the affected population is learners without quiet, secure or stable study settings, and the immediate evidence concerns space, movement, contact and material safety. In comparing informal settlements and unstable housing, a valid account states who was exposed, what institutional response was offered, who could use it and which learner condition was later observed. Within evidence on informal settlements and unstable housing, these stages should not collapse into a single indicator.[REF-03] [REF-04] [REF-39] [REF-55]

The principal risk is that urban residence is assumed to mean adequate remote access. In comparing informal settlements and unstable housing, that error can make excluded learners invisible or assign unequal effects to personal characteristics rather than institutions and material conditions. Within evidence on informal settlements and unstable housing, review should reconstruct entry into each source, actual service receipt and the barrier controlled by households, schools or public authorities.[REF-03] [REF-04] [REF-39]

The recommended response is to identify within-city housing and service barriers. Within evidence on informal settlements and unstable housing, reporting should include counts, group levels, reference period, source and material missingness. For interpretation of informal settlements and unstable housing, a gap or ratio should appear beside absolute conditions so apparent parity does not conceal low provision for all. In decisions about informal settlements and unstable housing, where the denominator or category is unstable, the conclusion should remain bounded.[REF-04] [REF-39] [REF-55]

Evidence for informal settlements and unstable housing should distinguish administrative status, household access, actual contact, educational use and later outcome. For interpretation of informal settlements and unstable housing, surveys, registers, provider records and experience evidence observe different populations. In decisions about informal settlements and unstable housing, agreement can strengthen a finding, while disagreement may reveal excluded settings, timing or definition differences that require explanation.[REF-03] [REF-55]

Intersectional interpretation of informal settlements and unstable housing should select characteristics tied to a plausible barrier and public decision. In decisions about informal settlements and unstable housing, disability, gender, location and income can combine with age, language or migration. For correction of informal settlements and unstable housing, counts, intervals and disclosure risk should determine the precision of publication. For informal settlements and unstable housing, small groups require neither invisibility nor unwarranted certainty.[REF-04] [REF-39] [REF-55]

Missing evidence for informal settlements and unstable housing is itself potentially unequal. For correction of informal settlements and unstable housing, learners unable to access education media may also be unable to answer connected surveys or submit work. For informal settlements and unstable housing, non-response should therefore be investigated by route and group. In comparing informal settlements and unstable housing, imputation or complete-case analysis should state assumptions and should not erase the likely direction of exclusion.[REF-03] [REF-04] [REF-39] [REF-55]

Causal interpretation of informal settlements and unstable housing should preserve baseline inequality, additional disruption, response differences and later change. For informal settlements and unstable housing, a post-disruption gap does not by itself identify its cause. In comparing informal settlements and unstable housing, evidence of exposure, timing and alternative explanations supports a bounded contribution claim; otherwise the report should describe association and the practical barrier requiring correction.[REF-04] [REF-39] [REF-55]

Public accountability for informal settlements and unstable housing should name the competent owner, resource, accessible correction and review date. In comparing informal settlements and unstable housing, learners need safe complaint and continuity routes. Within evidence on informal settlements and unstable housing, later review should ask whether the corrected service reached the group and whether participation, learning or wellbeing improved, while retaining the original definition and evidence limitations.[REF-03] [REF-04] [REF-39] [REF-55]

Camps, displacement and jurisdiction defines a particular comparison within the displacement geography. For camps, displacement and jurisdiction, the affected population is refugee and displaced learners across temporary provision and national systems, and the immediate evidence concerns provider, location, recognition and continuity. In comparing camps, displacement and jurisdiction, a valid account states who was exposed, what institutional response was offered, who could use it and which learner condition was later observed. Within evidence on camps, displacement and jurisdiction, these stages should not collapse into a single indicator.[REF-31] [REF-37] [REF-41] [REF-52]

The principal risk is that learners fall between response plans or records. In comparing camps, displacement and jurisdiction, that error can make excluded learners invisible or assign unequal effects to personal characteristics rather than institutions and material conditions. Within evidence on camps, displacement and jurisdiction, review should reconstruct entry into each source, actual service receipt and the barrier controlled by households, schools or public authorities.[REF-31] [REF-37] [REF-41]

The recommended response is to assign jurisdiction and preserve education continuity. Within evidence on camps, displacement and jurisdiction, reporting should include counts, group levels, reference period, source and material missingness. For interpretation of camps, displacement and jurisdiction, a gap or ratio should appear beside absolute conditions so apparent parity does not conceal low provision for all. In decisions about camps, displacement and jurisdiction, where the denominator or category is unstable, the conclusion should remain bounded.[REF-37] [REF-41] [REF-52]

Evidence for camps, displacement and jurisdiction should distinguish administrative status, household access, actual contact, educational use and later outcome. For interpretation of camps, displacement and jurisdiction, surveys, registers, provider records and experience evidence observe different populations. In decisions about camps, displacement and jurisdiction, agreement can strengthen a finding, while disagreement may reveal excluded settings, timing or definition differences that require explanation.[REF-31] [REF-52]

Intersectional interpretation of camps, displacement and jurisdiction should select characteristics tied to a plausible barrier and public decision. In decisions about camps, displacement and jurisdiction, disability, gender, location and income can combine with age, language or migration. For correction of camps, displacement and jurisdiction, counts, intervals and disclosure risk should determine the precision of publication. For camps, displacement and jurisdiction, small groups require neither invisibility nor unwarranted certainty.[REF-37] [REF-41] [REF-52]

Missing evidence for camps, displacement and jurisdiction is itself potentially unequal. For correction of camps, displacement and jurisdiction, learners unable to access education media may also be unable to answer connected surveys or submit work. For camps, displacement and jurisdiction, non-response should therefore be investigated by route and group. In comparing camps, displacement and jurisdiction, imputation or complete-case analysis should state assumptions and should not erase the likely direction of exclusion.[REF-31] [REF-37] [REF-41] [REF-52]

Causal interpretation of camps, displacement and jurisdiction should preserve baseline inequality, additional disruption, response differences and later change. For camps, displacement and jurisdiction, a post-disruption gap does not by itself identify its cause. In comparing camps, displacement and jurisdiction, evidence of exposure, timing and alternative explanations supports a bounded contribution claim; otherwise the report should describe association and the practical barrier requiring correction.[REF-37] [REF-41] [REF-52]

Public accountability for camps, displacement and jurisdiction should name the competent owner, resource, accessible correction and review date. In comparing camps, displacement and jurisdiction, learners need safe complaint and continuity routes. Within evidence on camps, displacement and jurisdiction, later review should ask whether the corrected service reached the group and whether participation, learning or wellbeing improved, while retaining the original definition and evidence limitations.[REF-31] [REF-37] [REF-41] [REF-52]

Return travel and reachable reopening defines a particular comparison within the reopening geography. For return travel and reachable reopening, the affected population is learners travelling to reopened institutions, and the immediate evidence concerns distance, transport, safety and attendance. In comparing return travel and reachable reopening, a valid account states who was exposed, what institutional response was offered, who could use it and which learner condition was later observed. Within evidence on return travel and reachable reopening, these stages should not collapse into a single indicator.[REF-11] [REF-32] [REF-50] [REF-56]

The principal risk is that formal reopening is counted despite inaccessible travel. In comparing return travel and reachable reopening, that error can make excluded learners invisible or assign unequal effects to personal characteristics rather than institutions and material conditions. Within evidence on return travel and reachable reopening, review should reconstruct entry into each source, actual service receipt and the barrier controlled by households, schools or public authorities.[REF-11] [REF-32] [REF-50]

The recommended response is to verify safe routes and actual return. Within evidence on return travel and reachable reopening, reporting should include counts, group levels, reference period, source and material missingness. For interpretation of return travel and reachable reopening, a gap or ratio should appear beside absolute conditions so apparent parity does not conceal low provision for all. In decisions about return travel and reachable reopening, where the denominator or category is unstable, the conclusion should remain bounded.[REF-32] [REF-50] [REF-56]

Evidence for return travel and reachable reopening should distinguish administrative status, household access, actual contact, educational use and later outcome. For interpretation of return travel and reachable reopening, surveys, registers, provider records and experience evidence observe different populations. In decisions about return travel and reachable reopening, agreement can strengthen a finding, while disagreement may reveal excluded settings, timing or definition differences that require explanation.[REF-11] [REF-56]

Intersectional interpretation of return travel and reachable reopening should select characteristics tied to a plausible barrier and public decision. In decisions about return travel and reachable reopening, disability, gender, location and income can combine with age, language or migration. For correction of return travel and reachable reopening, counts, intervals and disclosure risk should determine the precision of publication. For return travel and reachable reopening, small groups require neither invisibility nor unwarranted certainty.[REF-32] [REF-50] [REF-56]

Missing evidence for return travel and reachable reopening is itself potentially unequal. For correction of return travel and reachable reopening, learners unable to access education media may also be unable to answer connected surveys or submit work. For return travel and reachable reopening, non-response should therefore be investigated by route and group. In comparing return travel and reachable reopening, imputation or complete-case analysis should state assumptions and should not erase the likely direction of exclusion.[REF-11] [REF-32] [REF-50] [REF-56]

Causal interpretation of return travel and reachable reopening should preserve baseline inequality, additional disruption, response differences and later change. For return travel and reachable reopening, a post-disruption gap does not by itself identify its cause. In comparing return travel and reachable reopening, evidence of exposure, timing and alternative explanations supports a bounded contribution claim; otherwise the report should describe association and the practical barrier requiring correction.[REF-32] [REF-50] [REF-56]

Public accountability for return travel and reachable reopening should name the competent owner, resource, accessible correction and review date. In comparing return travel and reachable reopening, learners need safe complaint and continuity routes. Within evidence on return travel and reachable reopening, later review should ask whether the corrected service reached the group and whether participation, learning or wellbeing improved, while retaining the original definition and evidence limitations.[REF-11] [REF-32] [REF-50] [REF-56]

Devices, data and recurring education cost defines a particular comparison within the direct cost. For devices, data and recurring education cost, the affected population is households financing remote participation, and the immediate evidence concerns purchase, data, electricity and repair. In comparing devices, data and recurring education cost, a valid account states who was exposed, what institutional response was offered, who could use it and which learner condition was later observed. Within evidence on devices, data and recurring education cost, these stages should not collapse into a single indicator.[REF-03] [REF-07] [REF-49] [REF-55]

The principal risk is that free content is described as free education despite household costs. In comparing devices, data and recurring education cost, that error can make excluded learners invisible or assign unequal effects to personal characteristics rather than institutions and material conditions. Within evidence on devices, data and recurring education cost, review should reconstruct entry into each source, actual service receipt and the barrier controlled by households, schools or public authorities.[REF-03] [REF-07] [REF-49]

The recommended response is to report recurring costs and provide public support. Within evidence on devices, data and recurring education cost, reporting should include counts, group levels, reference period, source and material missingness. For interpretation of devices, data and recurring education cost, a gap or ratio should appear beside absolute conditions so apparent parity does not conceal low provision for all. In decisions about devices, data and recurring education cost, where the denominator or category is unstable, the conclusion should remain bounded.[REF-07] [REF-49] [REF-55]

Evidence for devices, data and recurring education cost should distinguish administrative status, household access, actual contact, educational use and later outcome. For interpretation of devices, data and recurring education cost, surveys, registers, provider records and experience evidence observe different populations. In decisions about devices, data and recurring education cost, agreement can strengthen a finding, while disagreement may reveal excluded settings, timing or definition differences that require explanation.[REF-03] [REF-55]

Intersectional interpretation of devices, data and recurring education cost should select characteristics tied to a plausible barrier and public decision. In decisions about devices, data and recurring education cost, disability, gender, location and income can combine with age, language or migration. For correction of devices, data and recurring education cost, counts, intervals and disclosure risk should determine the precision of publication. For devices, data and recurring education cost, small groups require neither invisibility nor unwarranted certainty.[REF-07] [REF-49] [REF-55]

Missing evidence for devices, data and recurring education cost is itself potentially unequal. For correction of devices, data and recurring education cost, learners unable to access education media may also be unable to answer connected surveys or submit work. For devices, data and recurring education cost, non-response should therefore be investigated by route and group. In comparing devices, data and recurring education cost, imputation or complete-case analysis should state assumptions and should not erase the likely direction of exclusion.[REF-03] [REF-07] [REF-49] [REF-55]

Causal interpretation of devices, data and recurring education cost should preserve baseline inequality, additional disruption, response differences and later change. For devices, data and recurring education cost, a post-disruption gap does not by itself identify its cause. In comparing devices, data and recurring education cost, evidence of exposure, timing and alternative explanations supports a bounded contribution claim; otherwise the report should describe association and the practical barrier requiring correction.[REF-07] [REF-49] [REF-55]

Public accountability for devices, data and recurring education cost should name the competent owner, resource, accessible correction and review date. In comparing devices, data and recurring education cost, learners need safe complaint and continuity routes. Within evidence on devices, data and recurring education cost, later review should ask whether the corrected service reached the group and whether participation, learning or wellbeing improved, while retaining the original definition and evidence limitations.[REF-03] [REF-07] [REF-49] [REF-55]

Study space and environmental conditions defines a particular comparison within the home environment. For study space and environmental conditions, the affected population is learners in crowded or insecure housing, and the immediate evidence concerns space, noise, safety and schedule. In comparing study space and environmental conditions, a valid account states who was exposed, what institutional response was offered, who could use it and which learner condition was later observed. Within evidence on study space and environmental conditions, these stages should not collapse into a single indicator.[REF-04] [REF-12] [REF-38] [REF-55]

The principal risk is that access measures ignore whether learning can occur. In comparing study space and environmental conditions, that error can make excluded learners invisible or assign unequal effects to personal characteristics rather than institutions and material conditions. Within evidence on study space and environmental conditions, review should reconstruct entry into each source, actual service receipt and the barrier controlled by households, schools or public authorities.[REF-04] [REF-12] [REF-38]

The recommended response is to include feasible environmental evidence and flexible routes. Within evidence on study space and environmental conditions, reporting should include counts, group levels, reference period, source and material missingness. For interpretation of study space and environmental conditions, a gap or ratio should appear beside absolute conditions so apparent parity does not conceal low provision for all. In decisions about study space and environmental conditions, where the denominator or category is unstable, the conclusion should remain bounded.[REF-12] [REF-38] [REF-55]

Evidence for study space and environmental conditions should distinguish administrative status, household access, actual contact, educational use and later outcome. For interpretation of study space and environmental conditions, surveys, registers, provider records and experience evidence observe different populations. In decisions about study space and environmental conditions, agreement can strengthen a finding, while disagreement may reveal excluded settings, timing or definition differences that require explanation.[REF-04] [REF-55]

Intersectional interpretation of study space and environmental conditions should select characteristics tied to a plausible barrier and public decision. In decisions about study space and environmental conditions, disability, gender, location and income can combine with age, language or migration. For correction of study space and environmental conditions, counts, intervals and disclosure risk should determine the precision of publication. For study space and environmental conditions, small groups require neither invisibility nor unwarranted certainty.[REF-12] [REF-38] [REF-55]

Missing evidence for study space and environmental conditions is itself potentially unequal. For correction of study space and environmental conditions, learners unable to access education media may also be unable to answer connected surveys or submit work. For study space and environmental conditions, non-response should therefore be investigated by route and group. In comparing study space and environmental conditions, imputation or complete-case analysis should state assumptions and should not erase the likely direction of exclusion.[REF-04] [REF-12] [REF-38] [REF-55]

Causal interpretation of study space and environmental conditions should preserve baseline inequality, additional disruption, response differences and later change. For study space and environmental conditions, a post-disruption gap does not by itself identify its cause. In comparing study space and environmental conditions, evidence of exposure, timing and alternative explanations supports a bounded contribution claim; otherwise the report should describe association and the practical barrier requiring correction.[REF-12] [REF-38] [REF-55]

Public accountability for study space and environmental conditions should name the competent owner, resource, accessible correction and review date. In comparing study space and environmental conditions, learners need safe complaint and continuity routes. Within evidence on study space and environmental conditions, later review should ask whether the corrected service reached the group and whether participation, learning or wellbeing improved, while retaining the original definition and evidence limitations.[REF-04] [REF-12] [REF-38] [REF-55]

Food security and school-linked support defines a particular comparison within the material wellbeing. For food security and school-linked support, the affected population is learners relying on school meals and social services, and the immediate evidence concerns replacement service and educational participation. In comparing food security and school-linked support, a valid account states who was exposed, what institutional response was offered, who could use it and which learner condition was later observed. Within evidence on food security and school-linked support, these stages should not collapse into a single indicator.[REF-03] [REF-11] [REF-50] [REF-52]

The principal risk is that learning measures ignore lost nutrition and support. In comparing food security and school-linked support, that error can make excluded learners invisible or assign unequal effects to personal characteristics rather than institutions and material conditions. Within evidence on food security and school-linked support, review should reconstruct entry into each source, actual service receipt and the barrier controlled by households, schools or public authorities.[REF-03] [REF-11] [REF-50]

The recommended response is to track alternative delivery and reach. Within evidence on food security and school-linked support, reporting should include counts, group levels, reference period, source and material missingness. For interpretation of food security and school-linked support, a gap or ratio should appear beside absolute conditions so apparent parity does not conceal low provision for all. In decisions about food security and school-linked support, where the denominator or category is unstable, the conclusion should remain bounded.[REF-11] [REF-50] [REF-52]

Evidence for food security and school-linked support should distinguish administrative status, household access, actual contact, educational use and later outcome. For interpretation of food security and school-linked support, surveys, registers, provider records and experience evidence observe different populations. In decisions about food security and school-linked support, agreement can strengthen a finding, while disagreement may reveal excluded settings, timing or definition differences that require explanation.[REF-03] [REF-52]

Intersectional interpretation of food security and school-linked support should select characteristics tied to a plausible barrier and public decision. In decisions about food security and school-linked support, disability, gender, location and income can combine with age, language or migration. For correction of food security and school-linked support, counts, intervals and disclosure risk should determine the precision of publication. For food security and school-linked support, small groups require neither invisibility nor unwarranted certainty.[REF-11] [REF-50] [REF-52]

Missing evidence for food security and school-linked support is itself potentially unequal. For correction of food security and school-linked support, learners unable to access education media may also be unable to answer connected surveys or submit work. For food security and school-linked support, non-response should therefore be investigated by route and group. In comparing food security and school-linked support, imputation or complete-case analysis should state assumptions and should not erase the likely direction of exclusion.[REF-03] [REF-11] [REF-50] [REF-52]

Causal interpretation of food security and school-linked support should preserve baseline inequality, additional disruption, response differences and later change. For food security and school-linked support, a post-disruption gap does not by itself identify its cause. In comparing food security and school-linked support, evidence of exposure, timing and alternative explanations supports a bounded contribution claim; otherwise the report should describe association and the practical barrier requiring correction.[REF-11] [REF-50] [REF-52]

Public accountability for food security and school-linked support should name the competent owner, resource, accessible correction and review date. In comparing food security and school-linked support, learners need safe complaint and continuity routes. Within evidence on food security and school-linked support, later review should ask whether the corrected service reached the group and whether participation, learning or wellbeing improved, while retaining the original definition and evidence limitations.[REF-03] [REF-11] [REF-50] [REF-52]

Adult education, time and instructional burden defines a particular comparison within the household support. For adult education, time and instructional burden, the affected population is learners requiring assistance with remote tasks, and the immediate evidence concerns caregiver availability, literacy and institutional contact. In comparing adult education, time and instructional burden, a valid account states who was exposed, what institutional response was offered, who could use it and which learner condition was later observed. Within evidence on adult education, time and instructional burden, these stages should not collapse into a single indicator.[REF-04] [REF-12] [REF-49] [REF-53]

The principal risk is that family support is assumed equal or families are expected to teach. In comparing adult education, time and instructional burden, that error can make excluded learners invisible or assign unequal effects to personal characteristics rather than institutions and material conditions. Within evidence on adult education, time and instructional burden, review should reconstruct entry into each source, actual service receipt and the barrier controlled by households, schools or public authorities.[REF-04] [REF-12] [REF-49]

The recommended response is to provide clear teacher guidance and low-burden tasks. Within evidence on adult education, time and instructional burden, reporting should include counts, group levels, reference period, source and material missingness. For interpretation of adult education, time and instructional burden, a gap or ratio should appear beside absolute conditions so apparent parity does not conceal low provision for all. In decisions about adult education, time and instructional burden, where the denominator or category is unstable, the conclusion should remain bounded.[REF-12] [REF-49] [REF-53]

Evidence for adult education, time and instructional burden should distinguish administrative status, household access, actual contact, educational use and later outcome. For interpretation of adult education, time and instructional burden, surveys, registers, provider records and experience evidence observe different populations. In decisions about adult education, time and instructional burden, agreement can strengthen a finding, while disagreement may reveal excluded settings, timing or definition differences that require explanation.[REF-04] [REF-53]

Intersectional interpretation of adult education, time and instructional burden should select characteristics tied to a plausible barrier and public decision. In decisions about adult education, time and instructional burden, disability, gender, location and income can combine with age, language or migration. For correction of adult education, time and instructional burden, counts, intervals and disclosure risk should determine the precision of publication. For adult education, time and instructional burden, small groups require neither invisibility nor unwarranted certainty.[REF-12] [REF-49] [REF-53]

Missing evidence for adult education, time and instructional burden is itself potentially unequal. For correction of adult education, time and instructional burden, learners unable to access education media may also be unable to answer connected surveys or submit work. For adult education, time and instructional burden, non-response should therefore be investigated by route and group. In comparing adult education, time and instructional burden, imputation or complete-case analysis should state assumptions and should not erase the likely direction of exclusion.[REF-04] [REF-12] [REF-49] [REF-53]

Causal interpretation of adult education, time and instructional burden should preserve baseline inequality, additional disruption, response differences and later change. For adult education, time and instructional burden, a post-disruption gap does not by itself identify its cause. In comparing adult education, time and instructional burden, evidence of exposure, timing and alternative explanations supports a bounded contribution claim; otherwise the report should describe association and the practical barrier requiring correction.[REF-12] [REF-49] [REF-53]

Public accountability for adult education, time and instructional burden should name the competent owner, resource, accessible correction and review date. In comparing adult education, time and instructional burden, learners need safe complaint and continuity routes. Within evidence on adult education, time and instructional burden, later review should ask whether the corrected service reached the group and whether participation, learning or wellbeing improved, while retaining the original definition and evidence limitations.[REF-04] [REF-12] [REF-49] [REF-53]

Income loss, work and participation defines a particular comparison within the economic shock. For income loss, work and participation, the affected population is learners entering paid work or household enterprise, and the immediate evidence concerns work hours, contact and continuing education. In comparing income loss, work and participation, a valid account states who was exposed, what institutional response was offered, who could use it and which learner condition was later observed. Within evidence on income loss, work and participation, these stages should not collapse into a single indicator.[REF-03] [REF-23] [REF-52] [REF-55]

The principal risk is that non-response is treated as disengagement rather than economic constraint. In comparing income loss, work and participation, that error can make excluded learners invisible or assign unequal effects to personal characteristics rather than institutions and material conditions. Within evidence on income loss, work and participation, review should reconstruct entry into each source, actual service receipt and the barrier controlled by households, schools or public authorities.[REF-03] [REF-23] [REF-52]

The recommended response is to investigate barriers and protect flexible continuation. Within evidence on income loss, work and participation, reporting should include counts, group levels, reference period, source and material missingness. For interpretation of income loss, work and participation, a gap or ratio should appear beside absolute conditions so apparent parity does not conceal low provision for all. In decisions about income loss, work and participation, where the denominator or category is unstable, the conclusion should remain bounded.[REF-23] [REF-52] [REF-55]

Evidence for income loss, work and participation should distinguish administrative status, household access, actual contact, educational use and later outcome. For interpretation of income loss, work and participation, surveys, registers, provider records and experience evidence observe different populations. In decisions about income loss, work and participation, agreement can strengthen a finding, while disagreement may reveal excluded settings, timing or definition differences that require explanation.[REF-03] [REF-55]

Intersectional interpretation of income loss, work and participation should select characteristics tied to a plausible barrier and public decision. In decisions about income loss, work and participation, disability, gender, location and income can combine with age, language or migration. For correction of income loss, work and participation, counts, intervals and disclosure risk should determine the precision of publication. For income loss, work and participation, small groups require neither invisibility nor unwarranted certainty.[REF-23] [REF-52] [REF-55]

Missing evidence for income loss, work and participation is itself potentially unequal. For correction of income loss, work and participation, learners unable to access education media may also be unable to answer connected surveys or submit work. For income loss, work and participation, non-response should therefore be investigated by route and group. In comparing income loss, work and participation, imputation or complete-case analysis should state assumptions and should not erase the likely direction of exclusion.[REF-03] [REF-23] [REF-52] [REF-55]

Causal interpretation of income loss, work and participation should preserve baseline inequality, additional disruption, response differences and later change. For income loss, work and participation, a post-disruption gap does not by itself identify its cause. In comparing income loss, work and participation, evidence of exposure, timing and alternative explanations supports a bounded contribution claim; otherwise the report should describe association and the practical barrier requiring correction.[REF-23] [REF-52] [REF-55]

Public accountability for income loss, work and participation should name the competent owner, resource, accessible correction and review date. In comparing income loss, work and participation, learners need safe complaint and continuity routes. Within evidence on income loss, work and participation, later review should ask whether the corrected service reached the group and whether participation, learning or wellbeing improved, while retaining the original definition and evidence limitations.[REF-03] [REF-23] [REF-52] [REF-55]

Relative wealth and changing hardship defines a particular comparison within the measurement caution. For relative wealth and changing hardship, the affected population is households changing economic position during disruption, and the immediate evidence concerns wealth grouping, current hardship and education access. In comparing relative wealth and changing hardship, a valid account states who was exposed, what institutional response was offered, who could use it and which learner condition was later observed. Within evidence on relative wealth and changing hardship, these stages should not collapse into a single indicator.[REF-07] [REF-08] [REF-38] [REF-55]

The principal risk is that pre-disruption wealth groups fully represent current resources. In comparing relative wealth and changing hardship, that error can make excluded learners invisible or assign unequal effects to personal characteristics rather than institutions and material conditions. Within evidence on relative wealth and changing hardship, review should reconstruct entry into each source, actual service receipt and the barrier controlled by households, schools or public authorities.[REF-07] [REF-08] [REF-38]

The recommended response is to state reference date and add direct access evidence. Within evidence on relative wealth and changing hardship, reporting should include counts, group levels, reference period, source and material missingness. For interpretation of relative wealth and changing hardship, a gap or ratio should appear beside absolute conditions so apparent parity does not conceal low provision for all. In decisions about relative wealth and changing hardship, where the denominator or category is unstable, the conclusion should remain bounded.[REF-08] [REF-38] [REF-55]

Evidence for relative wealth and changing hardship should distinguish administrative status, household access, actual contact, educational use and later outcome. For interpretation of relative wealth and changing hardship, surveys, registers, provider records and experience evidence observe different populations. In decisions about relative wealth and changing hardship, agreement can strengthen a finding, while disagreement may reveal excluded settings, timing or definition differences that require explanation.[REF-07] [REF-55]

Intersectional interpretation of relative wealth and changing hardship should select characteristics tied to a plausible barrier and public decision. In decisions about relative wealth and changing hardship, disability, gender, location and income can combine with age, language or migration. For correction of relative wealth and changing hardship, counts, intervals and disclosure risk should determine the precision of publication. For relative wealth and changing hardship, small groups require neither invisibility nor unwarranted certainty.[REF-08] [REF-38] [REF-55]

Missing evidence for relative wealth and changing hardship is itself potentially unequal. For correction of relative wealth and changing hardship, learners unable to access education media may also be unable to answer connected surveys or submit work. For relative wealth and changing hardship, non-response should therefore be investigated by route and group. In comparing relative wealth and changing hardship, imputation or complete-case analysis should state assumptions and should not erase the likely direction of exclusion.[REF-07] [REF-08] [REF-38] [REF-55]

Causal interpretation of relative wealth and changing hardship should preserve baseline inequality, additional disruption, response differences and later change. For relative wealth and changing hardship, a post-disruption gap does not by itself identify its cause. In comparing relative wealth and changing hardship, evidence of exposure, timing and alternative explanations supports a bounded contribution claim; otherwise the report should describe association and the practical barrier requiring correction.[REF-08] [REF-38] [REF-55]

Public accountability for relative wealth and changing hardship should name the competent owner, resource, accessible correction and review date. In comparing relative wealth and changing hardship, learners need safe complaint and continuity routes. Within evidence on relative wealth and changing hardship, later review should ask whether the corrected service reached the group and whether participation, learning or wellbeing improved, while retaining the original definition and evidence limitations.[REF-07] [REF-08] [REF-38] [REF-55]

Selecting decision-relevant intersections defines a particular comparison within the intersection choice. For selecting decision-relevant intersections, the affected population is learners affected by more than one material condition, and the immediate evidence concerns joint exposure, service and outcome. In comparing selecting decision-relevant intersections, a valid account states who was exposed, what institutional response was offered, who could use it and which learner condition was later observed. Within evidence on selecting decision-relevant intersections, these stages should not collapse into a single indicator.[REF-08] [REF-09] [REF-38] [REF-53]

The principal risk is that every possible cross-tabulation is attempted or no intersection is examined. In comparing selecting decision-relevant intersections, that error can make excluded learners invisible or assign unequal effects to personal characteristics rather than institutions and material conditions. Within evidence on selecting decision-relevant intersections, review should reconstruct entry into each source, actual service receipt and the barrier controlled by households, schools or public authorities.[REF-08] [REF-09] [REF-38]

The recommended response is to select combinations tied to plausible barriers and action. Within evidence on selecting decision-relevant intersections, reporting should include counts, group levels, reference period, source and material missingness. For interpretation of selecting decision-relevant intersections, a gap or ratio should appear beside absolute conditions so apparent parity does not conceal low provision for all. In decisions about selecting decision-relevant intersections, where the denominator or category is unstable, the conclusion should remain bounded.[REF-09] [REF-38] [REF-53]

Evidence for selecting decision-relevant intersections should distinguish administrative status, household access, actual contact, educational use and later outcome. For interpretation of selecting decision-relevant intersections, surveys, registers, provider records and experience evidence observe different populations. In decisions about selecting decision-relevant intersections, agreement can strengthen a finding, while disagreement may reveal excluded settings, timing or definition differences that require explanation.[REF-08] [REF-53]

Intersectional interpretation of selecting decision-relevant intersections should select characteristics tied to a plausible barrier and public decision. In decisions about selecting decision-relevant intersections, disability, gender, location and income can combine with age, language or migration. For correction of selecting decision-relevant intersections, counts, intervals and disclosure risk should determine the precision of publication. For selecting decision-relevant intersections, small groups require neither invisibility nor unwarranted certainty.[REF-09] [REF-38] [REF-53]

Missing evidence for selecting decision-relevant intersections is itself potentially unequal. For correction of selecting decision-relevant intersections, learners unable to access education media may also be unable to answer connected surveys or submit work. For selecting decision-relevant intersections, non-response should therefore be investigated by route and group. In comparing selecting decision-relevant intersections, imputation or complete-case analysis should state assumptions and should not erase the likely direction of exclusion.[REF-08] [REF-09] [REF-38] [REF-53]

Causal interpretation of selecting decision-relevant intersections should preserve baseline inequality, additional disruption, response differences and later change. For selecting decision-relevant intersections, a post-disruption gap does not by itself identify its cause. In comparing selecting decision-relevant intersections, evidence of exposure, timing and alternative explanations supports a bounded contribution claim; otherwise the report should describe association and the practical barrier requiring correction.[REF-09] [REF-38] [REF-53]

Public accountability for selecting decision-relevant intersections should name the competent owner, resource, accessible correction and review date. In comparing selecting decision-relevant intersections, learners need safe complaint and continuity routes. Within evidence on selecting decision-relevant intersections, later review should ask whether the corrected service reached the group and whether participation, learning or wellbeing improved, while retaining the original definition and evidence limitations.[REF-08] [REF-09] [REF-38] [REF-53]

Counts, rates and small denominators defines a particular comparison within the small-group evidence. For counts, rates and small denominators, the affected population is small populations represented in surveys or records, and the immediate evidence concerns count, rate and interval. In comparing counts, rates and small denominators, a valid account states who was exposed, what institutional response was offered, who could use it and which learner condition was later observed. Within evidence on counts, rates and small denominators, these stages should not collapse into a single indicator.[REF-05] [REF-08] [REF-19] [REF-38]

The principal risk is that large percentages from tiny groups drive firm conclusions. In comparing counts, rates and small denominators, that error can make excluded learners invisible or assign unequal effects to personal characteristics rather than institutions and material conditions. Within evidence on counts, rates and small denominators, review should reconstruct entry into each source, actual service receipt and the barrier controlled by households, schools or public authorities.[REF-05] [REF-08] [REF-19]

The recommended response is to publish denominators and bounded language. Within evidence on counts, rates and small denominators, reporting should include counts, group levels, reference period, source and material missingness. For interpretation of counts, rates and small denominators, a gap or ratio should appear beside absolute conditions so apparent parity does not conceal low provision for all. In decisions about counts, rates and small denominators, where the denominator or category is unstable, the conclusion should remain bounded.[REF-08] [REF-19] [REF-38]

Evidence for counts, rates and small denominators should distinguish administrative status, household access, actual contact, educational use and later outcome. For interpretation of counts, rates and small denominators, surveys, registers, provider records and experience evidence observe different populations. In decisions about counts, rates and small denominators, agreement can strengthen a finding, while disagreement may reveal excluded settings, timing or definition differences that require explanation.[REF-05] [REF-38]

Intersectional interpretation of counts, rates and small denominators should select characteristics tied to a plausible barrier and public decision. In decisions about counts, rates and small denominators, disability, gender, location and income can combine with age, language or migration. For correction of counts, rates and small denominators, counts, intervals and disclosure risk should determine the precision of publication. For counts, rates and small denominators, small groups require neither invisibility nor unwarranted certainty.[REF-08] [REF-19] [REF-38]

Missing evidence for counts, rates and small denominators is itself potentially unequal. For correction of counts, rates and small denominators, learners unable to access education media may also be unable to answer connected surveys or submit work. For counts, rates and small denominators, non-response should therefore be investigated by route and group. In comparing counts, rates and small denominators, imputation or complete-case analysis should state assumptions and should not erase the likely direction of exclusion.[REF-05] [REF-08] [REF-19] [REF-38]

Causal interpretation of counts, rates and small denominators should preserve baseline inequality, additional disruption, response differences and later change. For counts, rates and small denominators, a post-disruption gap does not by itself identify its cause. In comparing counts, rates and small denominators, evidence of exposure, timing and alternative explanations supports a bounded contribution claim; otherwise the report should describe association and the practical barrier requiring correction.[REF-08] [REF-19] [REF-38]

Public accountability for counts, rates and small denominators should name the competent owner, resource, accessible correction and review date. In comparing counts, rates and small denominators, learners need safe complaint and continuity routes. Within evidence on counts, rates and small denominators, later review should ask whether the corrected service reached the group and whether participation, learning or wellbeing improved, while retaining the original definition and evidence limitations.[REF-05] [REF-08] [REF-19] [REF-38]

Differential non-response and invisibility defines a particular comparison within the missing-data mechanism. For differential non-response and invisibility, the affected population is learners least able to answer remote surveys or tasks, and the immediate evidence concerns response route, missingness and likely direction. In comparing differential non-response and invisibility, a valid account states who was exposed, what institutional response was offered, who could use it and which learner condition was later observed. Within evidence on differential non-response and invisibility, these stages should not collapse into a single indicator.[REF-07] [REF-09] [REF-20] [REF-55]

The principal risk is that complete cases are treated as representative. In comparing differential non-response and invisibility, that error can make excluded learners invisible or assign unequal effects to personal characteristics rather than institutions and material conditions. Within evidence on differential non-response and invisibility, review should reconstruct entry into each source, actual service receipt and the barrier controlled by households, schools or public authorities.[REF-07] [REF-09] [REF-20]

The recommended response is to investigate and report who could not respond. Within evidence on differential non-response and invisibility, reporting should include counts, group levels, reference period, source and material missingness. For interpretation of differential non-response and invisibility, a gap or ratio should appear beside absolute conditions so apparent parity does not conceal low provision for all. In decisions about differential non-response and invisibility, where the denominator or category is unstable, the conclusion should remain bounded.[REF-09] [REF-20] [REF-55]

Evidence for differential non-response and invisibility should distinguish administrative status, household access, actual contact, educational use and later outcome. For interpretation of differential non-response and invisibility, surveys, registers, provider records and experience evidence observe different populations. In decisions about differential non-response and invisibility, agreement can strengthen a finding, while disagreement may reveal excluded settings, timing or definition differences that require explanation.[REF-07] [REF-55]

Intersectional interpretation of differential non-response and invisibility should select characteristics tied to a plausible barrier and public decision. In decisions about differential non-response and invisibility, disability, gender, location and income can combine with age, language or migration. For correction of differential non-response and invisibility, counts, intervals and disclosure risk should determine the precision of publication. For differential non-response and invisibility, small groups require neither invisibility nor unwarranted certainty.[REF-09] [REF-20] [REF-55]

Missing evidence for differential non-response and invisibility is itself potentially unequal. For correction of differential non-response and invisibility, learners unable to access education media may also be unable to answer connected surveys or submit work. For differential non-response and invisibility, non-response should therefore be investigated by route and group. In comparing differential non-response and invisibility, imputation or complete-case analysis should state assumptions and should not erase the likely direction of exclusion.[REF-07] [REF-09] [REF-20] [REF-55]

Causal interpretation of differential non-response and invisibility should preserve baseline inequality, additional disruption, response differences and later change. For differential non-response and invisibility, a post-disruption gap does not by itself identify its cause. In comparing differential non-response and invisibility, evidence of exposure, timing and alternative explanations supports a bounded contribution claim; otherwise the report should describe association and the practical barrier requiring correction.[REF-09] [REF-20] [REF-55]

Public accountability for differential non-response and invisibility should name the competent owner, resource, accessible correction and review date. In comparing differential non-response and invisibility, learners need safe complaint and continuity routes. Within evidence on differential non-response and invisibility, later review should ask whether the corrected service reached the group and whether participation, learning or wellbeing improved, while retaining the original definition and evidence limitations.[REF-07] [REF-09] [REF-20] [REF-55]

Administrative and survey source alignment defines a particular comparison within the source comparison. For administrative and survey source alignment, the affected population is authorities combining enrolment, household and service records, and the immediate evidence concerns definition, period and coverage. In comparing administrative and survey source alignment, a valid account states who was exposed, what institutional response was offered, who could use it and which learner condition was later observed. Within evidence on administrative and survey source alignment, these stages should not collapse into a single indicator.[REF-05] [REF-19] [REF-25] [REF-56]

The principal risk is that unlike sources are merged without mapping. In comparing administrative and survey source alignment, that error can make excluded learners invisible or assign unequal effects to personal characteristics rather than institutions and material conditions. Within evidence on administrative and survey source alignment, review should reconstruct entry into each source, actual service receipt and the barrier controlled by households, schools or public authorities.[REF-05] [REF-19] [REF-25]

The recommended response is to retain source identity and reconcile categories. Within evidence on administrative and survey source alignment, reporting should include counts, group levels, reference period, source and material missingness. For interpretation of administrative and survey source alignment, a gap or ratio should appear beside absolute conditions so apparent parity does not conceal low provision for all. In decisions about administrative and survey source alignment, where the denominator or category is unstable, the conclusion should remain bounded.[REF-19] [REF-25] [REF-56]

Evidence for administrative and survey source alignment should distinguish administrative status, household access, actual contact, educational use and later outcome. For interpretation of administrative and survey source alignment, surveys, registers, provider records and experience evidence observe different populations. In decisions about administrative and survey source alignment, agreement can strengthen a finding, while disagreement may reveal excluded settings, timing or definition differences that require explanation.[REF-05] [REF-56]

Intersectional interpretation of administrative and survey source alignment should select characteristics tied to a plausible barrier and public decision. In decisions about administrative and survey source alignment, disability, gender, location and income can combine with age, language or migration. For correction of administrative and survey source alignment, counts, intervals and disclosure risk should determine the precision of publication. For administrative and survey source alignment, small groups require neither invisibility nor unwarranted certainty.[REF-19] [REF-25] [REF-56]

Missing evidence for administrative and survey source alignment is itself potentially unequal. For correction of administrative and survey source alignment, learners unable to access education media may also be unable to answer connected surveys or submit work. For administrative and survey source alignment, non-response should therefore be investigated by route and group. In comparing administrative and survey source alignment, imputation or complete-case analysis should state assumptions and should not erase the likely direction of exclusion.[REF-05] [REF-19] [REF-25] [REF-56]

Causal interpretation of administrative and survey source alignment should preserve baseline inequality, additional disruption, response differences and later change. For administrative and survey source alignment, a post-disruption gap does not by itself identify its cause. In comparing administrative and survey source alignment, evidence of exposure, timing and alternative explanations supports a bounded contribution claim; otherwise the report should describe association and the practical barrier requiring correction.[REF-19] [REF-25] [REF-56]

Public accountability for administrative and survey source alignment should name the competent owner, resource, accessible correction and review date. In comparing administrative and survey source alignment, learners need safe complaint and continuity routes. Within evidence on administrative and survey source alignment, later review should ask whether the corrected service reached the group and whether participation, learning or wellbeing improved, while retaining the original definition and evidence limitations.[REF-05] [REF-19] [REF-25] [REF-56]

Confidentiality under detailed reporting defines a particular comparison within the disclosure protection. For confidentiality under detailed reporting, the affected population is learners represented in small local or disability groups, and the immediate evidence concerns identity risk, access and publication. In comparing confidentiality under detailed reporting, a valid account states who was exposed, what institutional response was offered, who could use it and which learner condition was later observed. Within evidence on confidentiality under detailed reporting, these stages should not collapse into a single indicator.[REF-09] [REF-20] [REF-21] [REF-38]

The principal risk is that detail exposes individuals or suppression conceals all need. In comparing confidentiality under detailed reporting, that error can make excluded learners invisible or assign unequal effects to personal characteristics rather than institutions and material conditions. Within evidence on confidentiality under detailed reporting, review should reconstruct entry into each source, actual service receipt and the barrier controlled by households, schools or public authorities.[REF-09] [REF-20] [REF-21]

The recommended response is to use protected access and safe aggregate accounts. Within evidence on confidentiality under detailed reporting, reporting should include counts, group levels, reference period, source and material missingness. For interpretation of confidentiality under detailed reporting, a gap or ratio should appear beside absolute conditions so apparent parity does not conceal low provision for all. In decisions about confidentiality under detailed reporting, where the denominator or category is unstable, the conclusion should remain bounded.[REF-20] [REF-21] [REF-38]

Evidence for confidentiality under detailed reporting should distinguish administrative status, household access, actual contact, educational use and later outcome. For interpretation of confidentiality under detailed reporting, surveys, registers, provider records and experience evidence observe different populations. In decisions about confidentiality under detailed reporting, agreement can strengthen a finding, while disagreement may reveal excluded settings, timing or definition differences that require explanation.[REF-09] [REF-38]

Intersectional interpretation of confidentiality under detailed reporting should select characteristics tied to a plausible barrier and public decision. In decisions about confidentiality under detailed reporting, disability, gender, location and income can combine with age, language or migration. For correction of confidentiality under detailed reporting, counts, intervals and disclosure risk should determine the precision of publication. For confidentiality under detailed reporting, small groups require neither invisibility nor unwarranted certainty.[REF-20] [REF-21] [REF-38]

Missing evidence for confidentiality under detailed reporting is itself potentially unequal. For correction of confidentiality under detailed reporting, learners unable to access education media may also be unable to answer connected surveys or submit work. For confidentiality under detailed reporting, non-response should therefore be investigated by route and group. In comparing confidentiality under detailed reporting, imputation or complete-case analysis should state assumptions and should not erase the likely direction of exclusion.[REF-09] [REF-20] [REF-21] [REF-38]

Causal interpretation of confidentiality under detailed reporting should preserve baseline inequality, additional disruption, response differences and later change. For confidentiality under detailed reporting, a post-disruption gap does not by itself identify its cause. In comparing confidentiality under detailed reporting, evidence of exposure, timing and alternative explanations supports a bounded contribution claim; otherwise the report should describe association and the practical barrier requiring correction.[REF-20] [REF-21] [REF-38]

Public accountability for confidentiality under detailed reporting should name the competent owner, resource, accessible correction and review date. In comparing confidentiality under detailed reporting, learners need safe complaint and continuity routes. Within evidence on confidentiality under detailed reporting, later review should ask whether the corrected service reached the group and whether participation, learning or wellbeing improved, while retaining the original definition and evidence limitations.[REF-09] [REF-20] [REF-21] [REF-38]

Measurement change and false trends defines a particular comparison within the temporal comparability. For measurement change and false trends, the affected population is users comparing pre-disruption and disruption evidence, and the immediate evidence concerns method, population and period. In comparing measurement change and false trends, a valid account states who was exposed, what institutional response was offered, who could use it and which learner condition was later observed. Within evidence on measurement change and false trends, these stages should not collapse into a single indicator.[REF-05] [REF-19] [REF-38] [REF-52]

The principal risk is that new collection modes create apparent group change. In comparing measurement change and false trends, that error can make excluded learners invisible or assign unequal effects to personal characteristics rather than institutions and material conditions. Within evidence on measurement change and false trends, review should reconstruct entry into each source, actual service receipt and the barrier controlled by households, schools or public authorities.[REF-05] [REF-19] [REF-38]

The recommended response is to publish breaks and sensitivity to method. Within evidence on measurement change and false trends, reporting should include counts, group levels, reference period, source and material missingness. For interpretation of measurement change and false trends, a gap or ratio should appear beside absolute conditions so apparent parity does not conceal low provision for all. In decisions about measurement change and false trends, where the denominator or category is unstable, the conclusion should remain bounded.[REF-19] [REF-38] [REF-52]

Evidence for measurement change and false trends should distinguish administrative status, household access, actual contact, educational use and later outcome. For interpretation of measurement change and false trends, surveys, registers, provider records and experience evidence observe different populations. In decisions about measurement change and false trends, agreement can strengthen a finding, while disagreement may reveal excluded settings, timing or definition differences that require explanation.[REF-05] [REF-52]

Intersectional interpretation of measurement change and false trends should select characteristics tied to a plausible barrier and public decision. In decisions about measurement change and false trends, disability, gender, location and income can combine with age, language or migration. For correction of measurement change and false trends, counts, intervals and disclosure risk should determine the precision of publication. For measurement change and false trends, small groups require neither invisibility nor unwarranted certainty.[REF-19] [REF-38] [REF-52]

Missing evidence for measurement change and false trends is itself potentially unequal. For correction of measurement change and false trends, learners unable to access education media may also be unable to answer connected surveys or submit work. For measurement change and false trends, non-response should therefore be investigated by route and group. In comparing measurement change and false trends, imputation or complete-case analysis should state assumptions and should not erase the likely direction of exclusion.[REF-05] [REF-19] [REF-38] [REF-52]

Causal interpretation of measurement change and false trends should preserve baseline inequality, additional disruption, response differences and later change. For measurement change and false trends, a post-disruption gap does not by itself identify its cause. In comparing measurement change and false trends, evidence of exposure, timing and alternative explanations supports a bounded contribution claim; otherwise the report should describe association and the practical barrier requiring correction.[REF-19] [REF-38] [REF-52]

Public accountability for measurement change and false trends should name the competent owner, resource, accessible correction and review date. In comparing measurement change and false trends, learners need safe complaint and continuity routes. Within evidence on measurement change and false trends, later review should ask whether the corrected service reached the group and whether participation, learning or wellbeing improved, while retaining the original definition and evidence limitations.[REF-05] [REF-19] [REF-38] [REF-52]

Finding, barrier and competent owner defines a particular comparison within the accountability chain. For finding, barrier and competent owner, the affected population is institutions able to correct unequal effects, and the immediate evidence concerns finding, system cause and authority. In comparing finding, barrier and competent owner, a valid account states who was exposed, what institutional response was offered, who could use it and which learner condition was later observed. Within evidence on finding, barrier and competent owner, these stages should not collapse into a single indicator.[REF-18] [REF-27] [REF-52] [REF-53]

The principal risk is that group disparities are described without identifying the body controlling access. In comparing finding, barrier and competent owner, that error can make excluded learners invisible or assign unequal effects to personal characteristics rather than institutions and material conditions. Within evidence on finding, barrier and competent owner, review should reconstruct entry into each source, actual service receipt and the barrier controlled by households, schools or public authorities.[REF-18] [REF-27] [REF-52]

The recommended response is to assign each material barrier to a competent owner. Within evidence on finding, barrier and competent owner, reporting should include counts, group levels, reference period, source and material missingness. For interpretation of finding, barrier and competent owner, a gap or ratio should appear beside absolute conditions so apparent parity does not conceal low provision for all. In decisions about finding, barrier and competent owner, where the denominator or category is unstable, the conclusion should remain bounded.[REF-27] [REF-52] [REF-53]

Evidence for finding, barrier and competent owner should distinguish administrative status, household access, actual contact, educational use and later outcome. For interpretation of finding, barrier and competent owner, surveys, registers, provider records and experience evidence observe different populations. In decisions about finding, barrier and competent owner, agreement can strengthen a finding, while disagreement may reveal excluded settings, timing or definition differences that require explanation.[REF-18] [REF-53]

Intersectional interpretation of finding, barrier and competent owner should select characteristics tied to a plausible barrier and public decision. In decisions about finding, barrier and competent owner, disability, gender, location and income can combine with age, language or migration. For correction of finding, barrier and competent owner, counts, intervals and disclosure risk should determine the precision of publication. For finding, barrier and competent owner, small groups require neither invisibility nor unwarranted certainty.[REF-27] [REF-52] [REF-53]

Missing evidence for finding, barrier and competent owner is itself potentially unequal. For correction of finding, barrier and competent owner, learners unable to access education media may also be unable to answer connected surveys or submit work. For finding, barrier and competent owner, non-response should therefore be investigated by route and group. In comparing finding, barrier and competent owner, imputation or complete-case analysis should state assumptions and should not erase the likely direction of exclusion.[REF-18] [REF-27] [REF-52] [REF-53]

Causal interpretation of finding, barrier and competent owner should preserve baseline inequality, additional disruption, response differences and later change. For finding, barrier and competent owner, a post-disruption gap does not by itself identify its cause. In comparing finding, barrier and competent owner, evidence of exposure, timing and alternative explanations supports a bounded contribution claim; otherwise the report should describe association and the practical barrier requiring correction.[REF-27] [REF-52] [REF-53]

Public accountability for finding, barrier and competent owner should name the competent owner, resource, accessible correction and review date. In comparing finding, barrier and competent owner, learners need safe complaint and continuity routes. Within evidence on finding, barrier and competent owner, later review should ask whether the corrected service reached the group and whether participation, learning or wellbeing improved, while retaining the original definition and evidence limitations.[REF-18] [REF-27] [REF-52] [REF-53]

Resource allocation by verified need defines a particular comparison within the equity finance. For resource allocation by verified need, the affected population is schools and communities needing additional capacity, and the immediate evidence concerns need, allocation, delivery and beneficiary. In comparing resource allocation by verified need, a valid account states who was exposed, what institutional response was offered, who could use it and which learner condition was later observed. Within evidence on resource allocation by verified need, these stages should not collapse into a single indicator.[REF-03] [REF-28] [REF-38] [REF-53]

The principal risk is that resources follow visible participation rather than excluded populations. In comparing resource allocation by verified need, that error can make excluded learners invisible or assign unequal effects to personal characteristics rather than institutions and material conditions. Within evidence on resource allocation by verified need, review should reconstruct entry into each source, actual service receipt and the barrier controlled by households, schools or public authorities.[REF-03] [REF-28] [REF-38]

The recommended response is to use barrier and reach evidence for transparent allocation. Within evidence on resource allocation by verified need, reporting should include counts, group levels, reference period, source and material missingness. For interpretation of resource allocation by verified need, a gap or ratio should appear beside absolute conditions so apparent parity does not conceal low provision for all. In decisions about resource allocation by verified need, where the denominator or category is unstable, the conclusion should remain bounded.[REF-28] [REF-38] [REF-53]

Evidence for resource allocation by verified need should distinguish administrative status, household access, actual contact, educational use and later outcome. For interpretation of resource allocation by verified need, surveys, registers, provider records and experience evidence observe different populations. In decisions about resource allocation by verified need, agreement can strengthen a finding, while disagreement may reveal excluded settings, timing or definition differences that require explanation.[REF-03] [REF-53]

Intersectional interpretation of resource allocation by verified need should select characteristics tied to a plausible barrier and public decision. In decisions about resource allocation by verified need, disability, gender, location and income can combine with age, language or migration. For correction of resource allocation by verified need, counts, intervals and disclosure risk should determine the precision of publication. For resource allocation by verified need, small groups require neither invisibility nor unwarranted certainty.[REF-28] [REF-38] [REF-53]

Missing evidence for resource allocation by verified need is itself potentially unequal. For correction of resource allocation by verified need, learners unable to access education media may also be unable to answer connected surveys or submit work. For resource allocation by verified need, non-response should therefore be investigated by route and group. In comparing resource allocation by verified need, imputation or complete-case analysis should state assumptions and should not erase the likely direction of exclusion.[REF-03] [REF-28] [REF-38] [REF-53]

Causal interpretation of resource allocation by verified need should preserve baseline inequality, additional disruption, response differences and later change. For resource allocation by verified need, a post-disruption gap does not by itself identify its cause. In comparing resource allocation by verified need, evidence of exposure, timing and alternative explanations supports a bounded contribution claim; otherwise the report should describe association and the practical barrier requiring correction.[REF-28] [REF-38] [REF-53]

Public accountability for resource allocation by verified need should name the competent owner, resource, accessible correction and review date. In comparing resource allocation by verified need, learners need safe complaint and continuity routes. Within evidence on resource allocation by verified need, later review should ask whether the corrected service reached the group and whether participation, learning or wellbeing improved, while retaining the original definition and evidence limitations.[REF-03] [REF-28] [REF-38] [REF-53]

Accessible and multi-medium correction defines a particular comparison within the service correction. For accessible and multi-medium correction, the affected population is learners excluded by the main education route, and the immediate evidence concerns alternative medium, accommodation and teacher contact. In comparing accessible and multi-medium correction, a valid account states who was exposed, what institutional response was offered, who could use it and which learner condition was later observed. Within evidence on accessible and multi-medium correction, these stages should not collapse into a single indicator.[REF-10] [REF-15] [REF-49] [REF-54]

The principal risk is that an additional resource repeats the same access barrier. In comparing accessible and multi-medium correction, that error can make excluded learners invisible or assign unequal effects to personal characteristics rather than institutions and material conditions. Within evidence on accessible and multi-medium correction, review should reconstruct entry into each source, actual service receipt and the barrier controlled by households, schools or public authorities.[REF-10] [REF-15] [REF-49]

The recommended response is to provide a coherent usable route with feedback. Within evidence on accessible and multi-medium correction, reporting should include counts, group levels, reference period, source and material missingness. For interpretation of accessible and multi-medium correction, a gap or ratio should appear beside absolute conditions so apparent parity does not conceal low provision for all. In decisions about accessible and multi-medium correction, where the denominator or category is unstable, the conclusion should remain bounded.[REF-15] [REF-49] [REF-54]

Evidence for accessible and multi-medium correction should distinguish administrative status, household access, actual contact, educational use and later outcome. For interpretation of accessible and multi-medium correction, surveys, registers, provider records and experience evidence observe different populations. In decisions about accessible and multi-medium correction, agreement can strengthen a finding, while disagreement may reveal excluded settings, timing or definition differences that require explanation.[REF-10] [REF-54]

Intersectional interpretation of accessible and multi-medium correction should select characteristics tied to a plausible barrier and public decision. In decisions about accessible and multi-medium correction, disability, gender, location and income can combine with age, language or migration. For correction of accessible and multi-medium correction, counts, intervals and disclosure risk should determine the precision of publication. For accessible and multi-medium correction, small groups require neither invisibility nor unwarranted certainty.[REF-15] [REF-49] [REF-54]

Missing evidence for accessible and multi-medium correction is itself potentially unequal. For correction of accessible and multi-medium correction, learners unable to access education media may also be unable to answer connected surveys or submit work. For accessible and multi-medium correction, non-response should therefore be investigated by route and group. In comparing accessible and multi-medium correction, imputation or complete-case analysis should state assumptions and should not erase the likely direction of exclusion.[REF-10] [REF-15] [REF-49] [REF-54]

Causal interpretation of accessible and multi-medium correction should preserve baseline inequality, additional disruption, response differences and later change. For accessible and multi-medium correction, a post-disruption gap does not by itself identify its cause. In comparing accessible and multi-medium correction, evidence of exposure, timing and alternative explanations supports a bounded contribution claim; otherwise the report should describe association and the practical barrier requiring correction.[REF-15] [REF-49] [REF-54]

Public accountability for accessible and multi-medium correction should name the competent owner, resource, accessible correction and review date. In comparing accessible and multi-medium correction, learners need safe complaint and continuity routes. Within evidence on accessible and multi-medium correction, later review should ask whether the corrected service reached the group and whether participation, learning or wellbeing improved, while retaining the original definition and evidence limitations.[REF-10] [REF-15] [REF-49] [REF-54]

Complaint and rapid remedy defines a particular comparison within the practical correction. For complaint and rapid remedy, the affected population is learners and families experiencing exclusion or harm, and the immediate evidence concerns safe report, response and continuity. In comparing complaint and rapid remedy, a valid account states who was exposed, what institutional response was offered, who could use it and which learner condition was later observed. Within evidence on complaint and rapid remedy, these stages should not collapse into a single indicator.[REF-14] [REF-21] [REF-22] [REF-50]

The principal risk is that complaints require the inaccessible medium or await aggregate review. In comparing complaint and rapid remedy, that error can make excluded learners invisible or assign unequal effects to personal characteristics rather than institutions and material conditions. Within evidence on complaint and rapid remedy, review should reconstruct entry into each source, actual service receipt and the barrier controlled by households, schools or public authorities.[REF-14] [REF-21] [REF-22]

The recommended response is to offer several routes and rapid educational remedy. Within evidence on complaint and rapid remedy, reporting should include counts, group levels, reference period, source and material missingness. For interpretation of complaint and rapid remedy, a gap or ratio should appear beside absolute conditions so apparent parity does not conceal low provision for all. In decisions about complaint and rapid remedy, where the denominator or category is unstable, the conclusion should remain bounded.[REF-21] [REF-22] [REF-50]

Evidence for complaint and rapid remedy should distinguish administrative status, household access, actual contact, educational use and later outcome. For interpretation of complaint and rapid remedy, surveys, registers, provider records and experience evidence observe different populations. In decisions about complaint and rapid remedy, agreement can strengthen a finding, while disagreement may reveal excluded settings, timing or definition differences that require explanation.[REF-14] [REF-50]

Intersectional interpretation of complaint and rapid remedy should select characteristics tied to a plausible barrier and public decision. In decisions about complaint and rapid remedy, disability, gender, location and income can combine with age, language or migration. For correction of complaint and rapid remedy, counts, intervals and disclosure risk should determine the precision of publication. For complaint and rapid remedy, small groups require neither invisibility nor unwarranted certainty.[REF-21] [REF-22] [REF-50]

Missing evidence for complaint and rapid remedy is itself potentially unequal. For correction of complaint and rapid remedy, learners unable to access education media may also be unable to answer connected surveys or submit work. For complaint and rapid remedy, non-response should therefore be investigated by route and group. In comparing complaint and rapid remedy, imputation or complete-case analysis should state assumptions and should not erase the likely direction of exclusion.[REF-14] [REF-21] [REF-22] [REF-50]

Causal interpretation of complaint and rapid remedy should preserve baseline inequality, additional disruption, response differences and later change. For complaint and rapid remedy, a post-disruption gap does not by itself identify its cause. In comparing complaint and rapid remedy, evidence of exposure, timing and alternative explanations supports a bounded contribution claim; otherwise the report should describe association and the practical barrier requiring correction.[REF-21] [REF-22] [REF-50]

Public accountability for complaint and rapid remedy should name the competent owner, resource, accessible correction and review date. In comparing complaint and rapid remedy, learners need safe complaint and continuity routes. Within evidence on complaint and rapid remedy, later review should ask whether the corrected service reached the group and whether participation, learning or wellbeing improved, while retaining the original definition and evidence limitations.[REF-14] [REF-21] [REF-22] [REF-50]

Recovery without punitive assumptions defines a particular comparison within the restoration equity. For recovery without punitive assumptions, the affected population is learners returning with unequal exposure and loss, and the immediate evidence concerns re-engagement, support and broad curriculum. In comparing recovery without punitive assumptions, a valid account states who was exposed, what institutional response was offered, who could use it and which learner condition was later observed. Within evidence on recovery without punitive assumptions, these stages should not collapse into a single indicator.[REF-29] [REF-50] [REF-52] [REF-56]

The principal risk is that return decisions penalise learners for inaccessible provision. In comparing recovery without punitive assumptions, that error can make excluded learners invisible or assign unequal effects to personal characteristics rather than institutions and material conditions. Within evidence on recovery without punitive assumptions, review should reconstruct entry into each source, actual service receipt and the barrier controlled by households, schools or public authorities.[REF-29] [REF-50] [REF-52]

The recommended response is to diagnose need and fund additional teaching. Within evidence on recovery without punitive assumptions, reporting should include counts, group levels, reference period, source and material missingness. For interpretation of recovery without punitive assumptions, a gap or ratio should appear beside absolute conditions so apparent parity does not conceal low provision for all. In decisions about recovery without punitive assumptions, where the denominator or category is unstable, the conclusion should remain bounded.[REF-50] [REF-52] [REF-56]

Evidence for recovery without punitive assumptions should distinguish administrative status, household access, actual contact, educational use and later outcome. For interpretation of recovery without punitive assumptions, surveys, registers, provider records and experience evidence observe different populations. In decisions about recovery without punitive assumptions, agreement can strengthen a finding, while disagreement may reveal excluded settings, timing or definition differences that require explanation.[REF-29] [REF-56]

Intersectional interpretation of recovery without punitive assumptions should select characteristics tied to a plausible barrier and public decision. In decisions about recovery without punitive assumptions, disability, gender, location and income can combine with age, language or migration. For correction of recovery without punitive assumptions, counts, intervals and disclosure risk should determine the precision of publication. For recovery without punitive assumptions, small groups require neither invisibility nor unwarranted certainty.[REF-50] [REF-52] [REF-56]

Missing evidence for recovery without punitive assumptions is itself potentially unequal. For correction of recovery without punitive assumptions, learners unable to access education media may also be unable to answer connected surveys or submit work. For recovery without punitive assumptions, non-response should therefore be investigated by route and group. In comparing recovery without punitive assumptions, imputation or complete-case analysis should state assumptions and should not erase the likely direction of exclusion.[REF-29] [REF-50] [REF-52] [REF-56]

Causal interpretation of recovery without punitive assumptions should preserve baseline inequality, additional disruption, response differences and later change. For recovery without punitive assumptions, a post-disruption gap does not by itself identify its cause. In comparing recovery without punitive assumptions, evidence of exposure, timing and alternative explanations supports a bounded contribution claim; otherwise the report should describe association and the practical barrier requiring correction.[REF-50] [REF-52] [REF-56]

Public accountability for recovery without punitive assumptions should name the competent owner, resource, accessible correction and review date. In comparing recovery without punitive assumptions, learners need safe complaint and continuity routes. Within evidence on recovery without punitive assumptions, later review should ask whether the corrected service reached the group and whether participation, learning or wellbeing improved, while retaining the original definition and evidence limitations.[REF-29] [REF-50] [REF-52] [REF-56]

Public reporting with temporal restraint defines a particular comparison within the bounded accountability. For public reporting with temporal restraint, the affected population is communities evaluating unequal disruption effects, and the immediate evidence concerns evidence date, uncertainty and later review. In comparing public reporting with temporal restraint, a valid account states who was exposed, what institutional response was offered, who could use it and which learner condition was later observed. Within evidence on public reporting with temporal restraint, these stages should not collapse into a single indicator.[REF-44] [REF-52] [REF-53] [REF-56]

The principal risk is that later outcomes are read backwards or early evidence becomes permanent labels. In comparing public reporting with temporal restraint, that error can make excluded learners invisible or assign unequal effects to personal characteristics rather than institutions and material conditions. Within evidence on public reporting with temporal restraint, review should reconstruct entry into each source, actual service receipt and the barrier controlled by households, schools or public authorities.[REF-44] [REF-52] [REF-53]

The recommended response is to date claims and plan correction and remeasurement. Within evidence on public reporting with temporal restraint, reporting should include counts, group levels, reference period, source and material missingness. For interpretation of public reporting with temporal restraint, a gap or ratio should appear beside absolute conditions so apparent parity does not conceal low provision for all. In decisions about public reporting with temporal restraint, where the denominator or category is unstable, the conclusion should remain bounded.[REF-52] [REF-53] [REF-56]

Evidence for public reporting with temporal restraint should distinguish administrative status, household access, actual contact, educational use and later outcome. For interpretation of public reporting with temporal restraint, surveys, registers, provider records and experience evidence observe different populations. In decisions about public reporting with temporal restraint, agreement can strengthen a finding, while disagreement may reveal excluded settings, timing or definition differences that require explanation.[REF-44] [REF-56]

Intersectional interpretation of public reporting with temporal restraint should select characteristics tied to a plausible barrier and public decision. In decisions about public reporting with temporal restraint, disability, gender, location and income can combine with age, language or migration. For correction of public reporting with temporal restraint, counts, intervals and disclosure risk should determine the precision of publication. For public reporting with temporal restraint, small groups require neither invisibility nor unwarranted certainty.[REF-52] [REF-53] [REF-56]

Missing evidence for public reporting with temporal restraint is itself potentially unequal. For correction of public reporting with temporal restraint, learners unable to access education media may also be unable to answer connected surveys or submit work. For public reporting with temporal restraint, non-response should therefore be investigated by route and group. In comparing public reporting with temporal restraint, imputation or complete-case analysis should state assumptions and should not erase the likely direction of exclusion.[REF-44] [REF-52] [REF-53] [REF-56]

Causal interpretation of public reporting with temporal restraint should preserve baseline inequality, additional disruption, response differences and later change. For public reporting with temporal restraint, a post-disruption gap does not by itself identify its cause. In comparing public reporting with temporal restraint, evidence of exposure, timing and alternative explanations supports a bounded contribution claim; otherwise the report should describe association and the practical barrier requiring correction.[REF-52] [REF-53] [REF-56]

Public accountability for public reporting with temporal restraint should name the competent owner, resource, accessible correction and review date. In comparing public reporting with temporal restraint, learners need safe complaint and continuity routes. Within evidence on public reporting with temporal restraint, later review should ask whether the corrected service reached the group and whether participation, learning or wellbeing improved, while retaining the original definition and evidence limitations.[REF-44] [REF-52] [REF-53] [REF-56]

An inequality record for exposure, response and outcome are distinct should preserve the group definition, population count, exposure, institutional offer, usable access, service receipt, outcome and uncertainty. For interpretation of exposure, response and outcome are distinct, it should state which barrier is actionable and by whom. In decisions about exposure, response and outcome are distinct, this keeps comparison connected to correction rather than turning a temporary disruption effect into a permanent description of the group.[REF-05] [REF-38] [REF-52] [REF-56]

An inequality record for availability, access, receipt and use should preserve the group definition, population count, exposure, institutional offer, usable access, service receipt, outcome and uncertainty. For interpretation of availability, access, receipt and use, it should state which barrier is actionable and by whom. In decisions about availability, access, receipt and use, this keeps comparison connected to correction rather than turning a temporary disruption effect into a permanent description of the group.[REF-07] [REF-49] [REF-55] [REF-56]

An inequality record for participation, learning and wellbeing should preserve the group definition, population count, exposure, institutional offer, usable access, service receipt, outcome and uncertainty. For interpretation of participation, learning and wellbeing, it should state which barrier is actionable and by whom. In decisions about participation, learning and wellbeing, this keeps comparison connected to correction rather than turning a temporary disruption effect into a permanent description of the group.[REF-10] [REF-12] [REF-50] [REF-53]

An inequality record for absolute levels and group gaps should preserve the group definition, population count, exposure, institutional offer, usable access, service receipt, outcome and uncertainty. For interpretation of absolute levels and group gaps, it should state which barrier is actionable and by whom. In decisions about absolute levels and group gaps, this keeps comparison connected to correction rather than turning a temporary disruption effect into a permanent description of the group.[REF-03] [REF-10] [REF-38] [REF-44]

An inequality record for starting inequality and additional disruption should preserve the group definition, population count, exposure, institutional offer, usable access, service receipt, outcome and uncertainty. For interpretation of starting inequality and additional disruption, it should state which barrier is actionable and by whom. In decisions about starting inequality and additional disruption, this keeps comparison connected to correction rather than turning a temporary disruption effect into a permanent description of the group.[REF-04] [REF-29] [REF-38] [REF-53]

An inequality record for causal restraint under changing conditions should preserve the group definition, population count, exposure, institutional offer, usable access, service receipt, outcome and uncertainty. For interpretation of causal restraint under changing conditions, it should state which barrier is actionable and by whom. In decisions about causal restraint under changing conditions, this keeps comparison connected to correction rather than turning a temporary disruption effect into a permanent description of the group.[REF-09] [REF-19] [REF-38] [REF-52]

An inequality record for accessible media and learning materials should preserve the group definition, population count, exposure, institutional offer, usable access, service receipt, outcome and uncertainty. For interpretation of accessible media and learning materials, it should state which barrier is actionable and by whom. In decisions about accessible media and learning materials, this keeps comparison connected to correction rather than turning a temporary disruption effect into a permanent description of the group.[REF-10] [REF-11] [REF-15] [REF-49]

An inequality record for assistive resources and technical support should preserve the group definition, population count, exposure, institutional offer, usable access, service receipt, outcome and uncertainty. For interpretation of assistive resources and technical support, it should state which barrier is actionable and by whom. In decisions about assistive resources and technical support, this keeps comparison connected to correction rather than turning a temporary disruption effect into a permanent description of the group.[REF-11] [REF-15] [REF-53] [REF-54]

An inequality record for specialist and therapeutic service interruption should preserve the group definition, population count, exposure, institutional offer, usable access, service receipt, outcome and uncertainty. For interpretation of specialist and therapeutic service interruption, it should state which barrier is actionable and by whom. In decisions about specialist and therapeutic service interruption, this keeps comparison connected to correction rather than turning a temporary disruption effect into a permanent description of the group.[REF-10] [REF-15] [REF-50] [REF-53]

An inequality record for caregiver burden and institutional duty should preserve the group definition, population count, exposure, institutional offer, usable access, service receipt, outcome and uncertainty. For interpretation of caregiver burden and institutional duty, it should state which barrier is actionable and by whom. In decisions about caregiver burden and institutional duty, this keeps comparison connected to correction rather than turning a temporary disruption effect into a permanent description of the group.[REF-12] [REF-49] [REF-52] [REF-55]

An inequality record for accessible assessment and participation evidence should preserve the group definition, population count, exposure, institutional offer, usable access, service receipt, outcome and uncertainty. For interpretation of accessible assessment and participation evidence, it should state which barrier is actionable and by whom. In decisions about accessible assessment and participation evidence, this keeps comparison connected to correction rather than turning a temporary disruption effect into a permanent description of the group.[REF-09] [REF-15] [REF-17] [REF-38]

An inequality record for return, accommodation and continuity should preserve the group definition, population count, exposure, institutional offer, usable access, service receipt, outcome and uncertainty. For interpretation of return, accommodation and continuity, it should state which barrier is actionable and by whom. In decisions about return, accommodation and continuity, this keeps comparison connected to correction rather than turning a temporary disruption effect into a permanent description of the group.[REF-11] [REF-15] [REF-50] [REF-53]

An inequality record for device priority and household allocation should preserve the group definition, population count, exposure, institutional offer, usable access, service receipt, outcome and uncertainty. For interpretation of device priority and household allocation, it should state which barrier is actionable and by whom. In decisions about device priority and household allocation, this keeps comparison connected to correction rather than turning a temporary disruption effect into a permanent description of the group.[REF-07] [REF-38] [REF-55] [REF-56]

An inequality record for care, domestic work and study time should preserve the group definition, population count, exposure, institutional offer, usable access, service receipt, outcome and uncertainty. For interpretation of care, domestic work and study time, it should state which barrier is actionable and by whom. In decisions about care, domestic work and study time, this keeps comparison connected to correction rather than turning a temporary disruption effect into a permanent description of the group.[REF-03] [REF-12] [REF-52] [REF-53]

An inequality record for safety, harassment and harmful contact should preserve the group definition, population count, exposure, institutional offer, usable access, service receipt, outcome and uncertainty. For interpretation of safety, harassment and harmful contact, it should state which barrier is actionable and by whom. In decisions about safety, harassment and harmful contact, this keeps comparison connected to correction rather than turning a temporary disruption effect into a permanent description of the group.[REF-14] [REF-20] [REF-21] [REF-50]

An inequality record for economic pressure and early leaving should preserve the group definition, population count, exposure, institutional offer, usable access, service receipt, outcome and uncertainty. For interpretation of economic pressure and early leaving, it should state which barrier is actionable and by whom. In decisions about economic pressure and early leaving, this keeps comparison connected to correction rather than turning a temporary disruption effect into a permanent description of the group.[REF-03] [REF-23] [REF-35] [REF-52]

An inequality record for pregnancy, parenthood and re-entry should preserve the group definition, population count, exposure, institutional offer, usable access, service receipt, outcome and uncertainty. For interpretation of pregnancy, parenthood and re-entry, it should state which barrier is actionable and by whom. In decisions about pregnancy, parenthood and re-entry, this keeps comparison connected to correction rather than turning a temporary disruption effect into a permanent description of the group.[REF-10] [REF-14] [REF-22] [REF-53]