ICEQC-R-2005-01 — Education for Sustainable Development: Implications for National Quality Policy

Education, public awareness and training have formed part of the international sustainable-development programme since Agenda 21. Chapter 36 identified three connected areas of action: reorienting education towards sustainable development, increasing public awareness and promoting training. It treated basic education as the foundation for wider environmental and development learning, and called for relevant concepts to be integrated across educational programmes, informed by scientific evidence and local conditions. The chapter also recognised formal and non-formal education and the need for teacher, administrator and planner preparation. Its approach built on the principles developed at the 1977 Intergovernmental Conference on Environmental Education in Tbilisi, including attention to the environment in its full social, economic and natural context, continuity across the educational life course, interdisciplinarity, critical examination and participation in addressing actual problems.[REF-01] [REF-12]

The Johannesburg Plan of Implementation renewed this direction in 2002 and located it alongside commitments to universal primary education, gender equality and national development. The Plan did not isolate sustainable-development learning from the wider condition of education systems. It connected access, educational quality, skills, public awareness, research, information and institutional capacity. That relationship matters for national policy: the Decade begins within an existing field of obligations, not on an empty administrative landscape.[REF-02]

General Assembly resolution 57/254 proclaimed the period 2005–2014 as the United Nations Decade and designated UNESCO as lead agency. Resolution 59/237 subsequently confirmed that the Decade began on 1 January 2005, reaffirmed the relationship with universal primary education and gender equality, and encouraged governments to consider measures within their educational systems, strategies and, where appropriate, national development plans. It also called for completion and consideration of an international implementation scheme. At the date of this report, the October 2004 scheme remains a draft under the process identified by the General Assembly.[REF-03] [REF-05] [REF-07]

The legal and administrative effect of these instruments differs across jurisdictions. A General Assembly resolution does not itself determine national curriculum, allocate domestic expenditure or establish a justiciable education standard in every system. It nonetheless records a collective commitment and a clear expectation of national consideration. Governments remain responsible for deciding how the commitment is expressed through their laws, plans, administrative standards and budget processes. A credible response should therefore identify the domestic authority for action rather than relying on the international declaration alone.

The change in 2005 is consequently one of policy status and implementation responsibility. Sustainable-development learning is no longer only a theme promoted through isolated projects or environmental programmes. The Decade places it before national authorities as a system question. The proper response is not necessarily a new statute or a new agency. It is a reasoned determination of where existing public frameworks are sufficient, where they are inconsistent and where amendment is required.

National quality policy determines what education is expected to accomplish and what evidence is accepted as showing that the expectation has been met. If education for sustainable development is omitted from that framework, it is likely to remain dependent on temporary projects, individual teacher commitment or external funding. If it is included only as aspirational language, institutions may be asked to claim alignment without any corresponding change in curriculum opportunity, professional capacity or learner experience. Both outcomes weaken accountability.

The 2005 Education for All Global Monitoring Report provides a useful analytical starting point. It treats education quality as a relationship among learner characteristics, context, enabling inputs, teaching and learning, and outcomes. Cognitive development is central, but education also has purposes connected with values, citizenship, creative and emotional development. The framework resists two errors: treating resources as if they were outcomes, and treating a limited set of measured outcomes as if they represented the whole educational purpose.[REF-08]

Education for sustainable development intersects with each element of that framework. Learners bring different languages, livelihoods, health conditions, environmental exposures and prior knowledge. Context includes patterns of production and consumption, public institutions, cultural values, local ecosystems and development choices. Enabling inputs include teachers, learning materials, safe facilities, time and access to reliable information. Teaching and learning require enquiry, application, discussion and the examination of consequences across disciplines. Outcomes include knowledge and skills, but also the capacity to make reasoned judgements, participate and continue learning.

This breadth does not justify an unlimited standard. A quality framework must specify what is within scope and what evidence is proportionate. The public authority responsible for education cannot guarantee that every learner will adopt prescribed private behaviour or that education alone will achieve sustainable development. It can reasonably be expected to secure an educational opportunity: access to relevant and accurate knowledge; development of intellectual and practical capabilities; experience of fair and participatory learning; and exposure to the ethical and civic questions established in public educational aims. Quality policy should evaluate the conditions and educational results within institutional control, not claim direct responsibility for every social outcome.

Rights instruments reinforce this distinction. The Convention on the Rights of the Child recognises both the right to education and broad aims for education, including development of the child's abilities, respect for human rights and cultural identity, preparation for responsible life in a free society, and respect for the natural environment. The Committee's General Comment No. 1 explains that these aims concern content, educational processes, pedagogical methods and the environment in which education takes place. Sustainable-development education should therefore be designed as an aspect of a rights-consistent educational experience, not as a demand for ideological conformity.[REF-09] [REF-10]

Education systems commonly respond to new public priorities by inserting material into curriculum documents. That response is visible and administratively convenient, but it can leave the determinants of practice unchanged. Where the timetable, examination system, teacher preparation, learning materials and inspection criteria continue to reward established routines, a new curricular statement may have little effect. The implementation problem is thus one of alignment.

Alignment has a vertical dimension. National purpose must be translated into curriculum expectations, teacher preparation, institutional plans and classroom practice. Evidence must travel in the opposite direction so that school and learner experience informs subnational and national review. A break at any point can convert policy into formal compliance. Ministries may publish an expectation that is not reflected in textbooks; teacher institutions may prepare practitioners for methods that assessment systems penalise; local authorities may receive reporting requirements without the means to collect reliable data.

Alignment also has a horizontal dimension. Education policy must relate coherently to national plans concerning poverty reduction, employment, health, environment, science, agriculture, water, energy and public participation. This does not mean that those sectors may transfer their objectives wholesale into the curriculum. Educational authorities retain responsibility for age appropriateness, subject integrity, instructional time and the rights of learners. Cross-government coordination should identify legitimate educational contributions and protect schools from accumulating unprioritised demands.

A third dimension is temporal. Curriculum and professional change take time, while public declarations create pressure for immediate evidence of action. Authorities should distinguish commencement measures from mature outcomes. In the first phase, defensible evidence may consist of an approved mandate, a completed baseline, defined responsibilities, revised teacher-education requirements, prepared materials and evaluated pilots. Claims of changed learner capability require later and stronger evidence. Reporting systems that do not distinguish these stages invite overstatement.

The following table identifies the principal forms of misalignment that national quality policy should examine.

Source: ICEQC analysis based on Agenda 21, the Johannesburg Plan of Implementation, General Assembly resolution 59/237, the Dakar Framework for Action and the EFA quality framework.[REF-01] [REF-02] [REF-05] [REF-06] [REF-08]

The Decade concerns values as well as knowledge, but public education policy must address values with care. International instruments provide a legitimate basis in human dignity, equality, responsibility, participation, respect for cultural identity and the natural environment. They do not authorise compulsory adherence to a partisan programme or the suppression of reasoned disagreement about policy choices. Quality standards should require accuracy, balance appropriate to the educational context, the distinction between evidence and assertion, and opportunities for learners to examine competing consequences.

Sustainable development also encompasses contested decisions. Energy policy, land use, industrial development, consumption and public finance involve distributional effects and uncertainty. Curricula should not present such questions as if scientific evidence automatically determines a single political decision. Scientific and technical understanding can clarify conditions, risks and likely consequences; public choices additionally involve rights, priorities, costs and the distribution of benefit and harm. Education should develop the capacity to distinguish these forms of reasoning.

National authorities should avoid defining quality by personal behaviour that institutions cannot fairly require or verify. A learner's household circumstances may determine transport, energy or consumption choices. Holding learners accountable for outcomes beyond their control would be inequitable. Appropriate assessment focuses on knowledge, analysis, practical competence within the educational setting and reasoned decision-making. Institutional policies may govern operations within institutional responsibility, but they should be proportionate to resources and public duties.

The commitment does not diminish the priority of foundational learning. Literacy, numeracy and scientific understanding enable learners to interpret information, assess claims and participate in decisions. In systems where many learners do not complete primary school or do not master basic skills, sustainable-development policy must strengthen rather than compete with those capabilities. The quality question is how foundational learning becomes relevant and usable, not whether it should be displaced by additional themes.[REF-06] [REF-08]

Nor should national policy assume that formal schooling is the only setting. Agenda 21 and the Decade encompass public awareness, training and non-formal learning. Adult education, vocational preparation, community learning and public information have distinct audiences and purposes. Quality arrangements should recognise these forms without importing school-based standards indiscriminately. The common concern is clarity of purpose, accuracy, accessibility, capable educators and evidence that learning opportunities reach the intended population.

A national response should begin with an explicit statement of educational purpose. Without it, curriculum revisions risk becoming a catalogue of topics. The statement should identify the capability that public education intends to develop and the population entitled to that opportunity. It should be sufficiently stable to guide institutions, yet sufficiently bounded to permit local relevance and professional judgement.

The international sources support a purpose with four connected elements. The first is **understanding**: learners should encounter reliable knowledge about relationships among social conditions, economic activity and environmental systems. The second is **capability**: they should be able to inquire, use evidence, solve problems and consider consequences over time. The third is **responsibility**: they should understand the rights and interests of others, including future generations, without being required to profess a uniform view on every public question. The fourth is **participation**: they should be prepared to contribute to decisions in institutions, communities and public life.[REF-01] [REF-07] [REF-09] [REF-10]

These elements need to be expressed as an entitlement rather than an optional enrichment. If access depends on school wealth, teacher enthusiasm or a learner's chosen subject, national policy will reproduce inequality. Entitlement does not mean identical content or identical pedagogy. It means that every learner within the relevant phase has a defined opportunity to develop the stated capabilities through content appropriate to context and level.

The statement should also identify relationships with existing educational aims. Sustainable-development learning overlaps with citizenship, science, geography, health, technical and vocational learning, economics, history, languages, ethics and the arts. Mapping those relationships can reduce duplication and reveal omissions. It can also protect subject integrity: a science curriculum should maintain standards of scientific evidence, while civic and ethical dimensions are addressed through appropriate disciplines and interdisciplinary work.

Quality review should ask whether the entitlement is intelligible to institutions and the public. Broad formulations such as “awareness of sustainability” are difficult to plan, teach or assess. A stronger formulation specifies what learners should know, be able to do and have an opportunity to consider. National authorities need not prescribe every learning activity, but they should define the minimum opportunity clearly enough for omission to be detected.

Curriculum integration must be planned across phases. Repetition of general messages does not constitute progression. Younger learners may begin with observation, care, health, relationships and the immediate environment. Later stages can introduce systems, evidence, competing uses of resources, uncertainty, economic and social distribution, and the analysis of policy alternatives. Vocational and higher education require closer connection to occupational practice, research, professional responsibility and sector-specific consequences.

Progression has at least five dimensions:

1. **Scale**, from immediate settings to national, regional and global relationships.
2. **Time**, from present effects to longer-term consequences and intergenerational considerations.
3. **Evidence**, from direct observation to the interpretation of data, models and uncertainty.
4. **Agency**, from participation in bounded institutional decisions to analysis of complex public choices.
5. **Integration**, from recognition of separate environmental, social and economic factors to reasoned examination of their interaction.

A curriculum framework should state how these dimensions develop. The framework need not assign every element to a new subject. It should identify where responsibility lies, where interdisciplinary work is expected and how omissions are prevented. Schools should not be left to infer an entire progression from an aspirational preamble.

Curriculum content should be selected through a transparent process. Relevance to national and local conditions is necessary, but local relevance alone is insufficient. Learners also need understanding of relationships that extend beyond their immediate experience. Climate, trade, migration, public health and resource use cross borders. Conversely, global materials can become abstract or culturally remote if they do not connect with local language, knowledge and livelihood. A sound framework moves between local and wider scales.

The treatment of indigenous and local knowledge requires respect and analytical clarity. Agenda 21 recognises the contribution of local communities and knowledge systems. National policy should enable such knowledge to be represented by or with the communities concerned, avoid extraction without context, and distinguish knowledge claims, values and practices. The standard of respectful inclusion does not require uncritical acceptance of every claim; neither does scientific review justify dismissing knowledge because it is transmitted through a different tradition. The educational task is to examine sources, purposes, evidence and context fairly.

The capabilities identified by the Decade cannot be developed entirely through transmission of conclusions. Learners need opportunities to ask questions, interpret evidence, compare explanations, discuss consequences and undertake practical work. Participatory methods have a substantive purpose: they allow learners to practise the judgement and cooperation that the policy seeks to develop. They should not be adopted as ritual or as a replacement for explicit teaching.

Effective pedagogy combines structured subject instruction with enquiry and application. Learners require sufficient knowledge to investigate a problem responsibly. Open discussion without a knowledge base can reward confidence rather than understanding; factual instruction without application can leave knowledge inert. Quality policy should therefore avoid a false choice between teacher direction and learner activity. The appropriate balance varies by age, content and prior attainment.

Assessment exerts a strong influence on implementation. If examinations and progression decisions focus exclusively on recall, institutions will rationally allocate time to what is measured. National authorities should review whether assessment specifications permit learners to demonstrate application, interpretation and reasoned judgement. This can be achieved within established subjects through carefully designed tasks; it does not require a single cross-curricular examination.

Assessment validity must remain paramount. Attitudes are difficult to measure reliably and can intrude on conscience if treated as required personal beliefs. Self-reported behaviour is vulnerable to social desirability and differences in household circumstances. National systems should be cautious about high-stakes use of such measures. Knowledge, analysis of evidence, explanation of trade-offs, design of feasible responses and reflection on a bounded practical task provide stronger educational evidence.

Comparative assessment evidence available at the end of 2004 illustrates both the value and the limits of outcome measures. PISA 2003 provides internationally comparable evidence on mathematical, reading and scientific literacy, learner dispositions and the distribution of outcomes across social groups in participating systems. Its results can inform thinking about applied knowledge and equity. They cannot be treated as a global measure of education for sustainable development, nor do they cover the full age range, countries or purposes at issue.[REF-11]

The relationship between curriculum and assessment should be reviewed before new expectations are implemented at scale. Appropriate questions include whether learners encounter sufficiently rich material, whether tasks require the intended reasoning, whether teachers can judge responses consistently, and whether language or cultural assumptions create avoidable barriers. Pilot evidence should be published with limitations rather than converted prematurely into national performance claims.

The quality of implementation will depend principally on teachers, but responsibility for creating the conditions of implementation remains with the system. Teachers cannot be expected to supply missing curriculum design, acquire new subject knowledge privately, create all materials and reconcile contradictory assessment incentives without institutional support. Policy that places the entire burden on individual commitment is neither credible nor equitable.

Teacher policy should address initial preparation, continuing professional learning, leadership support, workload, access to materials and professional collaboration. Preparation needs vary. A primary teacher may require support to connect learning across subjects while maintaining depth. A secondary subject teacher may require methods for interdisciplinary planning without losing disciplinary standards. Vocational educators need current sector knowledge and the capacity to examine environmental and social implications of practice. Adult educators may need approaches suited to learners whose experience is the principal resource for discussion.

Professional learning should be tied to curriculum and assessment expectations. Short awareness sessions can introduce the Decade but are unlikely to change practice. More substantial development includes examination of subject content, planning of lessons and tasks, observation or review of learner work, and supported revision. Evidence of attendance alone should not be reported as evidence of capability or classroom implementation.

The distribution of prepared teachers requires attention. Rural, remote, disadvantaged and rapidly expanding systems frequently face the greatest shortages and the least access to continuing education. A universal policy implemented through voluntary or fee-based professional development may increase inequality. National plans should identify priority groups, delivery constraints and the additional cost of reaching them. Distance methods may widen access but require reliable infrastructure, appropriate materials and opportunities for professional interaction.

School leaders, inspectors, curriculum advisers and teacher educators also require preparation. If leaders treat the policy as an event programme, or if inspection continues to reward narrow coverage, teacher development will have limited effect. Quality policy should define the contribution expected from each role and provide consistent guidance. The aim is not to create a new specialist post in every institution, but to distribute responsibility through existing professional structures with sufficient expertise.

An education institution communicates through its organisation as well as its curriculum. Learners will observe how decisions are made, whose views are heard, how resources are used, whether facilities are safe and accessible, and whether stated commitments apply in practice. Institutional review should therefore examine the congruence between educational aims and the learning environment.

This principle must be applied proportionately. A poorly resourced school should not be penalised for infrastructure beyond its control. Review should distinguish institutional decisions from structural constraints and should direct evidence about the latter to the responsible authority. The purpose is to improve conditions and accountability, not to transfer public obligations to institutions with the least capacity.

Appropriate institutional evidence may include arrangements for learner participation, integration of local issues into learning, use of the school grounds or community as a learning resource, safety and health conditions, management of water and materials, and relationships with families and local bodies. These matters should be connected to educational purpose. Operational activity without learner engagement may improve institutional management but does not by itself demonstrate learning; classroom discussion without any opportunity to observe or participate may limit application.

Participation requires safeguards. Learners should have meaningful opportunities to contribute within age-appropriate boundaries, but they should not carry responsibility that belongs to public authorities or staff. Consultation should not expose children to conflict, unsafe work or pressure to represent an entire group. Institutions should explain which decisions are open to influence, how views will be considered and what constraints apply.

Community partnership can improve relevance and extend expertise. It can also introduce unequal influence, commercial interests or unverified material. Institutions need procedures for selecting partners, checking educational content, protecting learners and preserving public responsibility for curriculum. Partnership should supplement institutional capacity without determining access according to a community's wealth or external connections.

The General Assembly encourages governments to consider implementation within education systems, strategies and national development plans. This requires a clear locus of responsibility. A cross-sectoral theme can easily become everyone's concern and no authority's accountable duty. National arrangements should identify the body responsible for education policy, the bodies that contribute subject expertise, the authority that approves curriculum and assessment changes, and the means by which finance and implementation are reviewed.[REF-05]

Coordination should begin from defined decisions rather than permanent consultation without output. An inter-ministerial arrangement may be useful where it resolves questions of policy coherence, evidence or resources. It should not bypass educational expertise or create direct demands on institutions from multiple departments. One consolidated national expectation is preferable to parallel guidance on environment, health, agriculture, citizenship and employment that competes for the same instructional time.

Where education powers are devolved, national policy should distinguish common entitlement from subnational discretion. Central authorities may define minimum aims, data definitions and equity safeguards; subnational authorities may determine content, support arrangements and implementation sequence. Responsibilities for funding must follow responsibilities for delivery. An unfunded national expectation can produce formal compliance and widen territorial inequality.

The relationship with national development planning should be reciprocal. Education plans should consider the knowledge and capabilities required by national development priorities. Development plans should recognise the time, resources and institutional conditions required for educational change. Neither document should treat education as a communications channel for decisions made elsewhere.

Public reporting should identify the status of each measure. Adopted requirements, pilot programmes, guidance and proposed reforms are not equivalent. The same discipline applies to the international framework: the October 2004 implementation scheme should be described as a draft pending the processes identified by the General Assembly. Precision about status strengthens rather than weakens the credibility of national action.

Every new quality expectation has a resource implication, even where no new subject or institution is created. Curriculum development requires expertise and consultation. Teacher learning requires time, materials and facilitation. New assessment tasks require design and moderation. Monitoring requires data capacity. Institutional projects may require equipment, safe facilities or transport. Failure to cost these elements shifts the burden to local institutions and households.

The first financing principle is additionality of purpose rather than necessarily of budget line. Authorities should determine which changes can be achieved by revising existing activity, which require reallocation and which require additional resources. Reorientation may reduce duplication across programmes, but savings should not be assumed before they are demonstrated. A costed plan should disclose both direct expenditure and demands on professional time.

The second principle is protection of foundational provision. Agenda 21, the Johannesburg Plan and the Dakar commitments all connect sustainable-development education with basic education. In systems facing shortages of teachers, learning materials or safe classrooms, funds should not be diverted from essential provision to visible but peripheral projects. Integration should strengthen the relevance and quality of ordinary teaching. Capital projects or technology purchases require an educational rationale, a maintenance plan and evidence that disadvantaged learners will not be excluded.[REF-01] [REF-02] [REF-06]

The third principle is distributional review. Equal allocations do not produce equal capacity where costs and needs differ. Rural distance, language diversity, small-school scale, disability access, conflict and weak infrastructure can raise implementation costs. National budgets should identify whether resources reach the institutions expected to make the largest change. Aggregate expenditure growth is insufficient evidence.

External assistance can support development and testing, particularly in low-income systems, but national policy should avoid dependence on short project cycles. Donor-funded materials or training should conform to the national framework, disclose recurrent costs and include arrangements for continuity. Where external priorities determine which regions or subjects receive support, authorities should assess the resulting equity and coherence.

The Decade begins in the same year that the Dakar and Millennium commitments sought the elimination of gender disparity in primary and secondary education. The international record indicates that progress remains uneven and that quality cannot be inferred from enrolment alone. Sustainable-development policy must therefore examine who enters, remains in and benefits from education, as well as what content is offered.[REF-02] [REF-05] [REF-06] [REF-08]

Content may affect groups differently. Examples drawn exclusively from urban consumption can be remote from rural learners; materials that assume secure access to water or energy can obscure unequal conditions; technical language can exclude learners studying in a second language; participation activities can reproduce gendered expectations about who speaks and who performs practical work. Curriculum review should examine representation and accessibility without reducing learners to fixed group categories.

Opportunity indicators should be disaggregated wherever data quality permits. Sex and location are immediate priorities in many systems, but national contexts may require attention to income, language, disability, ethnicity, migration status or other dimensions. Disaggregation should be governed by rights, statistical quality and protection from harm. Where a group is absent from the data, the report should record the evidence gap rather than imply universal coverage.

Equity also concerns the distribution of environmental and development risks. Learners in poor communities may have more direct experience of unsafe water, pollution, land degradation, insecure livelihoods or disaster. Education should enable examination of those conditions without treating affected communities as objects of study or implying that local behaviour is the sole cause. National and wider structural factors belong within the analysis.

The standard of success is not identical activity in every institution. It is equitable access to the intended learning capability, supported by resources proportionate to need. National quality policy should permit contextual variation while requiring evidence that variation does not conceal a lower entitlement for disadvantaged learners.

The opening of the Decade creates a demand for evidence before a settled international measurement system exists. National authorities should neither postpone all monitoring until a comprehensive instrument is available nor create an elaborate index from weak data. A staged framework is required. It should establish a baseline, record implementation conditions, examine educational opportunity and introduce outcome measures only where their validity is sufficient for the use proposed.

Monitoring should answer distinct questions. **Commitment indicators** show whether the mandate has been incorporated into authorised policy. **Capacity indicators** show whether personnel, materials, time and finance are available. **Opportunity indicators** show whether learners encounter the intended curriculum and pedagogy. **Outcome indicators** show what learners know or can do. **Equity indicators** show how capacity, opportunity and outcomes are distributed. Combining these questions into a single count would obscure where implementation has failed.

The distinction between system, institution and learner evidence should also be preserved. A national curriculum provision is system evidence; it does not prove that every institution offers the provision. A school plan is institutional evidence; it does not prove that learners experienced or understood the planned activity. A learner assessment can show performance on specified tasks; it does not by itself explain the curriculum, resources or teaching that contributed to the result. Public reporting should present the chain rather than select whichever evidence produces the most favourable account.

Baseline evidence should be collected before major intervention. Where policy has already commenced, authorities should document the timing and identify which measures pre-date the baseline. The baseline need not cover every possible dimension. A small set of well-defined measures, repeated consistently and supported by qualitative investigation, will be more useful than a wide set collected once.

The framework should be designed for improvement as well as accountability. Schools and local authorities need evidence at a level and frequency that can inform decisions. National reporting needs sufficient consistency for distributional analysis and public oversight. These purposes may require different instruments. Requiring every institution to collect the full national research dataset would impose unnecessary burden and may reduce accuracy.

The following dimensions provide a defensible starting point for national monitoring in 2005. They draw on established quality concepts while recognising the early stage of policy implementation.

Source: ICEQC policy synthesis based on the EFA quality framework and the implementation responsibilities established in the cited instruments.[REF-05] [REF-08]

Every indicator should have a written specification. At minimum, the specification should identify its purpose, definition, population, numerator and denominator where relevant, data source, reference period, level of disaggregation, frequency, known limitations and responsible body. Changes in definition should be recorded and historical series should not be presented as continuous where comparability has been broken.

Activity counts require particular discipline. The number of teachers attending a workshop may be useful for delivery management. It does not establish that participants completed the programme, acquired the intended capability, applied it in their work or reached learners. Reporting should label the measure as participation and, where possible, connect it to subsequent evidence of practice.

Curriculum mapping should distinguish presence from depth. A keyword search can identify documents containing terms such as “environment” or “sustainability”, but it cannot show whether the treatment is accurate, progressive or assessed. Expert review should examine the nature of the learning expectation, the time and resources available and its relationship to other content.

Learner outcome measures should be introduced through a published validity programme. Tasks may ask learners to interpret data, identify relationships, explain uncertainty, compare options or design a response within stated constraints. Scoring criteria should reward the quality of reasoning and use of evidence, not conformity with a preferred policy conclusion. Language demand, cultural familiarity and access to relevant prior learning should be examined before comparisons are reported.

Evaluation should be matched to the maturity and scale of the intervention. A new curriculum framework may first require a process evaluation: whether institutions understood the expectation, whether materials arrived, whether teachers had adequate preparation and what obstacles arose. An outcome evaluation becomes appropriate when implementation is sufficiently established for learners to have received the intended opportunity.

Claims of causation require more than improvement over time. Learner outcomes may change because of wider curriculum reform, teacher recruitment, demographic shifts, economic conditions or changes in assessment. Where a comparison group or phased introduction is feasible and ethical, it may strengthen inference. Where it is not, evaluators should use multiple sources, examine plausible alternative explanations and state the limits of attribution.

Qualitative evidence is necessary in an emerging domain. Interviews, observation, work samples and institutional case studies can reveal how policy is interpreted and why implementation differs. Such evidence should be sampled and analysed systematically. Illustrative success stories should not be presented as representative national findings unless the selection method supports that conclusion.

Evaluation should record unintended effects. Possible effects include displacement of instructional time, additional teacher workload, superficial project activity, exclusion of learners unable to meet participation costs, politicisation of content, or concentration of resources in already capable institutions. Positive spillovers may include improved subject relevance, community engagement, teacher collaboration or better use of local evidence. Both require investigation rather than assumption.

National authorities should publish negative and mixed findings. The Decade's credibility will depend on the capacity to learn from implementation, not on uninterrupted claims of success. A pilot that identifies an unworkable assessment model or an inequitable delivery method has produced useful evidence if the finding informs revision.

Public reporting should separate inputs, activity, opportunity and outcomes. A concise national statement can present the mandate, implementation coverage, resource position, principal findings, distributional differences, limitations and next actions. Technical annexes should contain definitions and methods. This structure enables public understanding without sacrificing auditability.

Reporting should avoid unqualified statements that education for sustainable development has been “integrated” or “mainstreamed”. These terms can refer to policy language, curriculum coverage, teacher preparation, institutional practice or learner experience. The report should specify which condition has been evidenced and at what scale.

Claims of national coverage should use a denominator. Reporting that 5,000 schools participated has limited meaning without the number and type of eligible schools, their geographic distribution and the participation of different learner groups. Where data cover only a programme or sample, that boundary should remain visible in headings, tables and conclusions.

International comparison should be approached cautiously. National curricula, administrative structures, languages, assessment traditions and available data differ. Comparison is most useful when it examines policy design, indicator definitions and implementation conditions. A ranking based on an immature composite measure would create false precision and incentives for formal compliance.

Independent scrutiny can strengthen confidence, but the form should reflect national institutions. Statistical offices, parliamentary bodies, audit institutions, research institutions or public consultation may contribute different kinds of review. Independence should not be claimed solely because a report was prepared outside the responsible ministry; reviewers need access to evidence, a disclosed method and freedom to report limitations.

Many low-income systems enter the Decade while seeking to expand primary participation, recruit teachers, provide basic materials and reduce completion gaps. Sustainable-development education should be integrated into these priorities rather than established as a separate programme competing for scarce capacity. The most immediate contribution may lie in improving relevance, strengthening foundational learning through locally meaningful applications and protecting the quality of expansion.

Policy design should begin from institutional reality. Where classes are large, materials limited and teachers have restricted preparation, complex project requirements may be unworkable. This does not justify a lower educational purpose. It requires a delivery model based on feasible routines, shared materials, clear subject connections and sustained teacher support. Practical enquiry can use the immediate environment and community knowledge, provided that safety, accuracy and learner burden are addressed.

National frameworks should resist dependence on costly equipment. Some topics require scientific instruments or technical resources, but many core capabilities—observation, classification, interpretation of simple data, discussion of consequences and planning within constraints—can be developed with modest materials. Technology should be selected for an identified educational function and supported over its usable life.

Expansion can weaken quality when teacher recruitment and school construction do not keep pace with enrolment. The EFA monitoring evidence places teachers, instructional time, materials and school conditions at the centre of quality. ESD expectations introduced without regard to those constraints may become additional text in policy documents while learners continue to lack the conditions for basic learning.[REF-08]

External support should reinforce national systems. Separate training cadres, reporting formats and demonstration schools may produce visible activity but limited scale or continuity. Assistance should contribute to national curriculum development, teacher-education capacity, local-language materials and data systems. Recurrent costs and responsibilities after project completion should be agreed at the outset.

Middle-income systems may possess stronger administrative and professional infrastructure while facing large internal disparities. The policy opportunity is to connect ESD with curriculum reform, applied learning and institutional review. The risk is that national averages and well-resourced demonstration institutions conceal unequal access to the richer forms of learning.

Curriculum congestion is a common constraint. Authorities should map existing content and remove avoidable duplication before adding requirements. Interdisciplinary work should be planned around defined questions and subject contributions rather than broad themes for which no teacher or department has clear responsibility. Assessment reform is especially significant where high-stakes examinations govern institutional behaviour.

Urban and rural contexts may require different content and support. Urban learners may examine transport, air quality, waste, housing and consumption; rural learners may examine land, water, agriculture, biodiversity and access to services. These are not exclusive categories, and all learners need understanding beyond their locality. The quality standard should concern progression and capability, not uniform examples.

Private provision, where present, should not create a separate entitlement. National curriculum and learner-protection duties should apply according to domestic law. Public authorities should ensure that required evidence is proportionate and that monitoring does not depend solely on self-description. Commercial sponsorship of materials or activities requires disclosure and educational review.

Greater resources do not ensure coherent education for sustainable development. High-income systems may have extensive subject provision, assessment capacity and access to information while maintaining fragmented curricula, strong examination incentives and substantial inequalities by social background. The policy task may be less about introducing topics than about connecting knowledge, application and institutional responsibility.

Comparative assessment evidence shows that participating high-income systems differ considerably in both performance and equity. The presence of advanced infrastructure should not be treated as evidence that all learners can apply knowledge or that opportunity is evenly distributed. National monitoring should examine the relationship between socio-economic background, school conditions and access to demanding forms of enquiry.[REF-11]

Consumer behaviour and resource use may be prominent topics, but education should not place the weight of structural change on individual learners. Analysis should include production systems, public policy, infrastructure, markets and institutional decisions. Personal responsibility belongs within a wider account of collective authority and unequal capacity.

High-income systems also have a responsibility to examine the international effects of domestic choices and the conditions under which knowledge, technology and resources are shared. This should be approached through evidence and educationally appropriate analysis, not through moral assertion detached from curriculum content.

Small island developing States may face direct relationships among education, coastal resources, water, energy, transport, disaster exposure and economic dependence. Geographic dispersion can increase the cost of teacher support, materials and data collection. National scale may also allow closer coordination among public bodies and institutions, but small professional communities can limit specialist capacity.

Curriculum relevance is particularly significant. Locally grounded study can connect science, livelihoods, culture and public decision-making. It should not confine learners to local concerns or deny access to wider knowledge. Partnerships with regional institutions may support curriculum expertise, teacher education and assessment while preserving national authority over content.

Monitoring frameworks should account for small numbers. Percentage changes can be unstable, and disaggregation may risk identifying individuals. Qualitative evidence and multi-year reporting may be required. Comparisons with large systems should disclose scale and structural differences.

In settings affected by conflict or emergency, continuity, safety and access take precedence. Sustainable-development education may support social understanding, health, resource management and reconstruction, but it should not add burdens that compromise essential provision. Curriculum decisions must consider trauma, contested narratives, teacher security and the risk of exposing learners through public activity.

Participation should be voluntary, safe and bounded. Community projects that are appropriate in stable settings may be unsafe where movement, land or public expression carries risk. Institutional review should focus on conditions within control and on referral of structural needs to responsible authorities.

Reconstruction provides an opportunity to improve facilities, curriculum and governance, but rapid programmes can import materials poorly suited to language or context. National and local educators should participate in design. Evidence of feasibility should precede scale.

In federal or decentralised systems, the Decade can expose uncertainty about which level defines aims, funds implementation, prepares teachers and reports results. A written allocation of responsibility is necessary. National coordination may establish broad entitlement and comparable evidence; regional or local authorities may determine detailed curriculum and support.

Comparability should be limited to measures with sufficiently consistent definitions. A national report can present different regional approaches without ranking them, examining how context, authority and resources shape implementation. Shared minimum metadata will allow evidence to be interpreted while respecting legitimate variation.

Funding arrangements require close review. If subnational authorities carry new responsibilities without revenue or transfers, implementation will follow fiscal capacity. Equal entitlement requires either adequate own resources or an allocation mechanism responsive to need and cost.

The first phase should establish what the system is undertaking and where it begins. National authorities should issue a bounded mandate, identify the responsible body, map existing curriculum and policy, and commission a baseline of institutional capacity and educational opportunity. Consultation should include teachers, learners, communities, subject specialists and authorities responsible for relevant public sectors.

Mapping should identify both assets and contradictions. Existing science, citizenship, health, vocational and local learning may already contribute substantially. Assessment rules, teacher requirements or inspection criteria may inhibit integration. The purpose is not to relabel all existing education as ESD, but to determine where the intended capabilities are genuinely supported and where gaps remain.

The baseline should use available data where definitions are adequate and collect new evidence selectively. It should describe missingness and coverage. A national inventory of projects can inform the review, but project presence should not be treated as learner coverage.

Authorities should select a limited number of changes with a clear theory of action. A priority may concern curriculum progression, teacher education, assessment tasks, institutional self-review or support for disadvantaged regions. Selection should be based on the baseline, expected educational value, feasibility, equity and the capacity to evaluate.

Design should state the problem, the change proposed, the actors responsible, required resources, expected sequence and evidence of progress. The theory of action should identify assumptions. For example, providing materials will not improve learning unless teachers understand the content, have time to use it and encounter assessment arrangements consistent with the intended method.

Teacher and leadership support should precede or accompany new expectations. Materials should be tested for accuracy, language, age appropriateness and usability. Where external partners contribute, intellectual and commercial interests should be disclosed and national review maintained.

Pilots should include settings that reflect the range of national conditions, not only institutions already known for innovation. Selection should include variation in location, resources, language, school size and learner population where feasible. The purpose is to test implementation requirements and equity, not to demonstrate that a programme can succeed under favourable conditions.

Evaluation questions should be agreed before implementation. Data should include delivery fidelity, adaptations, resource use, teacher experience, learner opportunity, outcomes appropriate to the period and unintended effects. Findings should explain context and should not be generalised beyond the sample without justification.

Pilot institutions should be protected from excessive reporting. A coherent research design can collect sufficient evidence without asking practitioners to produce repeated narrative accounts for multiple bodies. Feedback should be returned to participating institutions in a form that supports improvement.

Scaling should follow evidence of feasibility and educational value. Revision may be required even where outcomes are promising if costs, workload or inequity make the model unsuitable for wider use. National scale may require simpler tools, stronger regional support or a longer timetable.

The decision to scale should publish the evidence considered, limitations, expected cost and measures to reach disadvantaged institutions. Requirements should be incorporated into ordinary curriculum, professional and accountability frameworks. Temporary project arrangements should not become permanent parallel systems unless there is a clear and reviewed reason.

Public reporting should continue through the stages of implementation. A first report may concern mandate and baseline; a later report may address coverage and capacity; outcome reporting should follow only when learners have had a credible opportunity to learn. This sequence protects the meaning of evidence.

On the evidence available at the commencement of the Decade, national authorities should consider the following actions:

1. establish a formal, publicly accessible statement connecting education for sustainable development with existing educational aims, rights and national plans;
2. preserve the draft status of the current international implementation scheme in official descriptions until the relevant process is complete;
3. define a minimum educational entitlement expressed through knowledge, capability, responsible judgement and participation, with progression across phases;
4. map curriculum, assessment, teacher policy, institutional review and finance before adding new requirements;
5. cost implementation and protect basic provision, with additional support directed to institutions facing greater need or delivery cost;
6. develop teacher, leader and adviser capacity through sustained, curriculum-linked professional learning rather than awareness activity alone;
7. review high-stakes assessment for consistency with application, enquiry and reasoned judgement;
8. establish a staged monitoring framework that separates commitment, capacity, opportunity, outcome and equity;
9. test policy in varied settings, publish limitations and revise before national scale; and
10. report publicly in terms that distinguish adopted policy, implementation coverage, educational opportunity and evidenced outcomes.

These considerations do not constitute a claim that one sequence will suit every jurisdiction. They identify the controls needed to move from an international commitment to a national quality policy capable of scrutiny and improvement.

The implications of the Decade begin before compulsory schooling. Early childhood provision influences language, health, social development, curiosity and children's relationship with their immediate surroundings. These foundations bear directly on later learning, but policy for this phase should not import abstract content or assessment models designed for older learners. The quality question is whether young children experience secure, inclusive and stimulating environments in which observation, care, cooperation and communication are developed through developmentally appropriate activity.

National policy should preserve the unity of care and education. Where responsibilities are divided among ministries or providers, sustainable-development language should not obscure unresolved matters of health, nutrition, safety, staffing and access. Environmental learning cannot be credible in an unsafe setting without water or sanitation. The policy contribution may therefore lie first in improving the learning environment and enabling children to explore it safely.

Curriculum guidance should favour direct experience, play, stories, language and guided observation. Concepts may arise through food, water, weather, plants, animals, family and community. The intention is not to assign responsibility for environmental or social problems to young children. It is to establish attentiveness, vocabulary, care and the capacity to ask and answer simple questions.

Practitioners require materials and professional guidance suited to the phase. A national curriculum statement addressed only to schools may leave early childhood providers without support. Where provision is diverse, minimum expectations should apply according to domestic authority, while support recognises differences in setting and staff preparation. Family and community participation can strengthen relevance but should respect the expertise and workload of carers and educators.

Monitoring at this stage should examine access, safety, staffing, materials and the quality of interaction. High-stakes testing of young children is not justified by the aims of the Decade. Observation and sampled review may inform improvement, provided that findings are not used to label children or institutions on the basis of narrow tasks. The EFA framework's attention to learner characteristics, enabling inputs and learning processes is particularly relevant.[REF-08]

Primary education carries two inseparable responsibilities: securing foundational capability and establishing the knowledge and habits through which learners can understand their world. Education for sustainable development should reinforce both. Reading, writing, numeracy and scientific observation are not preliminary matters to be completed before relevant learning begins; they are developed through well-sequenced encounters with meaningful content.

Curriculum design should identify where essential concepts enter and how they progress. Science may address living systems, materials, water and evidence. Geography may develop understanding of place, resources and human-environment relationships. Languages provide access to information, explanation and public communication. Mathematics enables measurement, comparison and representation. Social and civic learning can examine rules, cooperation and decisions. Integration should preserve the standards of each area.

The immediate environment provides accessible material, but access to it is unequal. Some schools have safe grounds, nearby services and community expertise; others operate in conditions where field activity is impracticable. National entitlement cannot depend on local amenity. Curriculum and resource design should provide alternative ways to achieve the same learning purpose, including shared images, maps, simple datasets, demonstrations and structured accounts from other settings.

Time allocation requires protection. Projects can consume substantial time while producing limited conceptual development. A quality standard should ask what learners are expected to know or do, how the activity contributes, and what evidence will show learning. Displays, clean-up events and commemorative days may support engagement, but participation is not a substitute for an instructional sequence.

Primary assessment should use language and tasks appropriate to learners' reading level and prior opportunity. Learners may classify evidence, explain a relationship, compare two options or record observations. Scoring should not reward access to household resources or adult assistance. National sample assessment may eventually examine selected capabilities, but classroom assessment and moderated work provide the first source of implementation evidence.

The Decade's commitment must be understood alongside universal primary completion. In systems where children remain excluded or leave early, an enriched curriculum reaches only those already served. Policy should report participation and completion with any account of new curriculum opportunity. It should also examine whether greater relevance supports attendance and persistence, without claiming causation unless the evaluation design permits it.[REF-02] [REF-05] [REF-06]

Secondary education permits greater disciplinary depth and more demanding analysis of interactions among economic activity, social conditions and environmental systems. It also introduces stronger subject boundaries and, in many systems, examinations that determine progression. National policy should address these structures directly rather than assume that a cross-curricular theme will organise itself.

Each subject should identify its distinctive contribution. Scientific disciplines address evidence, mechanisms, uncertainty and the testing of explanations. Geography examines spatial relationships, resources and human-environment interaction. History examines change, institutions and the consequences of past choices. Economics can address production, distribution, incentives and external costs. Mathematics supports modelling and critical interpretation of quantitative claims. Languages and the arts develop communication, perspective and cultural understanding. Civic and ethical study addresses rights, responsibilities and public reasoning.

Interdisciplinary learning becomes valuable when a question requires these contributions. It becomes weak when disciplinary knowledge is replaced by broad discussion. Schools need planning time, a defined question, agreed responsibilities and assessment arrangements. A common project should not require every subject to suspend its own progression or generate artificial links.

Upper-secondary pathways create an equity concern. Learners in academic, vocational and other programmes should receive an entitlement appropriate to their studies. Sustainable-development questions should not be reserved for high-status academic subjects, nor should vocational learners receive only behavioural instruction without access to underlying knowledge and public debate. Conversely, academic provision should connect theory with practical consequences.

Examinations are likely to determine the depth of implementation. National assessment bodies should review whether existing tasks already permit relevant application and whether new contexts can be introduced without compromising reliability. Changes should follow piloting, examiner preparation and publication of expectations. Sudden assessment reform can disadvantage cohorts and institutions with less capacity to adapt.

Career guidance should present emerging occupational and educational pathways accurately. It should not promise employment outcomes that evidence cannot support. Learners need information about qualifications, working conditions, further study and the changing knowledge requirements of sectors affected by environmental and development policy.

Technical and vocational education and training connects learning directly with production, services, technology and occupational practice. It therefore occupies a central place in the Decade. Quality policy should ensure that relevant environmental, social, safety and resource considerations are integrated into occupational competence rather than offered only through general awareness modules.

The content should be sector-specific. Agricultural learning may examine soil, water, productivity and local knowledge; construction may address materials, energy, safety and durability; tourism may examine cultural and environmental assets, employment and community effects; manufacturing may consider process efficiency, waste, worker protection and quality control. The applicable knowledge and responsibility differ, and national occupational standards or curricula should state them precisely.

Integration should not dilute core technical competence. A learner cannot evaluate resource efficiency or environmental risk without understanding the process concerned. Practical assessment should therefore examine safe and competent performance together with the ability to recognise relevant consequences and follow applicable procedures. Values statements detached from occupational evidence offer weak assurance.

Employer and worker participation can improve currency and relevance. Public authorities should preserve educational and public-interest objectives where commercial incentives differ. Consultation should include small enterprises and informal or community-based work where these form a substantial part of the economy, not only large employers with greater administrative capacity.

Teacher and trainer capability may be constrained by outdated equipment, limited industry contact or insufficient preparation in new requirements. Professional development should be linked to actual occupational change and accessible across regions. Equipment investment requires maintenance, safe use and a plan for equitable access. A high-cost demonstration centre does not establish national learner entitlement.

Qualifications and assessment arrangements should make relevant competence visible. Where evidence comes from workplace learning, responsibilities for supervision, authenticity and learner protection should be clear. Workplaces should not use sustainability-related projects to obtain unpaid labour or transfer regulatory duties to learners.

Higher education contributes through teaching, professional formation, research, institutional practice and public engagement. National quality policy should recognise these functions without prescribing a uniform curriculum across disciplines or compromising academic enquiry. The common expectation is that institutions identify the relevance of sustainable-development questions to their missions and provide credible evidence for the claims they make.

Professional programmes bear particular responsibilities. Engineering, teacher education, health, agriculture, planning, economics, public administration and law influence decisions with long-term social and environmental consequences. Programme design should examine whether graduates encounter the applicable knowledge, methods, professional duties and uncertainty. Inclusion should be based on disciplinary and occupational analysis rather than a general compulsory statement disconnected from practice.

Research policy should support inquiry across disciplines and protect standards of evidence. Sustainable development involves natural sciences, social sciences, humanities and professional fields. Interdisciplinary programmes require clear questions and methods; the presence of several disciplines does not by itself produce integrated knowledge. Public communication should distinguish established findings, uncertainty and policy judgement.

Institutional operations can provide a setting for teaching and research. Energy, procurement, transport, land, buildings and community relationships may generate evidence and practical questions. Operational targets should not be reported as educational outcomes unless learners are meaningfully involved and learning is assessed. Institutions should also avoid presenting partial operational improvements as proof of overall sustainability.

Access and participation remain part of quality. An institution may offer advanced study while excluding groups through cost, location, language or prior educational inequality. National reporting should examine who enters and completes relevant programmes, not only the number of programmes created.

Academic freedom and public accountability should be treated as compatible. Institutions and scholars need freedom to test evidence and disagree about policy. Public funding and public claims require transparent methods, responsible use of resources and disclosure of limitations. Quality policy should protect both conditions.

Agenda 21 and the Decade extend beyond formal institutions. Adults require learning related to livelihoods, health, public services, civic participation and changing technologies. Non-formal and community education can reach populations for whom school-based delivery is inaccessible or inappropriate. Its quality should be judged by purpose, relevance, accessibility, educator capability and evidence of learning rather than by resemblance to school.

Literacy is a central enabling capability. Written public information, technical guidance and participation processes exclude adults who cannot use them. Programmes should connect literacy and numeracy with meaningful contexts while ensuring that the curriculum develops transferable capability. A narrow focus on messages can leave participants dependent on others to interpret future information.

Adult learners bring substantial experience. Pedagogy should allow that experience to inform the learning process while providing access to reliable knowledge and opportunities to question established practice. Community knowledge should be respected, and disagreement should be handled without coercion. Participation in a programme should not be a condition for access to unrelated public benefits.

Provision may be delivered by public bodies, community organisations, employers or other providers. National arrangements should define minimum safeguards and evidence according to the authority and risk involved. Public information should identify the provider, purpose, cost, duration and any recognised outcome. Claims about employment, income or community effects require evidence.

Monitoring participation is difficult where provision is dispersed and intermittent. Counts should identify unique participants where possible, duration, completion definition and population coverage. Evaluation should include accessibility, retention, demonstrated capability and use of learning, while recognising that later social outcomes are influenced by many factors.

Teacher education is both a phase of higher education and the principal means of sustaining change across the system. The Decade will not be secured through repeated short courses if initial preparation continues unchanged. National authorities should review programme requirements, placement experience, teacher-educator capacity and the evidence required for completion.

Every teacher does not require identical expertise. All require an understanding of the national educational purpose, the ability to use evidence responsibly and the capacity to create an inclusive learning environment. Subject and phase specialisation should determine additional knowledge and method. Teacher educators need sufficient expertise to model the pedagogy and assessment expected in schools.

Placement institutions should provide opportunities to observe and practise relevant teaching without requiring trainees to introduce unsupported innovations. Mentors need guidance and time. Assessment of trainees should examine planning, subject accuracy, adaptation to learners and reflection on evidence. Personal advocacy should not substitute for professional competence.

Continuing professional learning should build on initial preparation and respond to curriculum change. Systems with large numbers of serving teachers require staged provision and regional capacity. Cascade models may extend reach but can lose depth at each stage; materials, trainer preparation and sampled review of delivery are necessary controls.

Workforce policy should consider retention. Additional expectations imposed in conditions of low pay, large classes or weak support can contribute to dissatisfaction and turnover. Sustainable implementation depends on the sustainability of the teaching profession itself. National plans should examine workload and opportunity cost, not only course participation.

Governments can act through law, regulation, curriculum, funding conditions, professional standards, guidance, institutional review and public reporting. The choice should reflect the authority required, the specificity of the expectation and the risk of non-implementation. Not every objective requires legislation; not every material duty should be left to non-binding guidance.

Law may establish broad educational aims, rights and public responsibilities. Curriculum frameworks can define entitlement and progression. Professional standards can identify capability. Funding instruments can support implementation and correct distributional inequality. Guidance can explain methods and examples while preserving professional judgement. Review and reporting can provide evidence and corrective action.

Instrument choice should avoid duplication and conflicting status. If a curriculum requirement is mandatory, supporting guidance should not present it as optional. If a pilot is voluntary, public statements should not imply national implementation. Authorities should maintain a register of current documents, their scope and their relationship to earlier instruments.

Sunset or review provisions are useful for new administrative requirements. A data return, grant condition or pilot standard should be reviewed after sufficient evidence has accumulated. Temporary arrangements should not persist solely because institutions have adapted to them.

Quality assurance can undermine quality when evidence demands consume the time needed for teaching and improvement. National authorities should require only evidence that serves a defined decision. Institutions should be permitted to use existing records where these provide reliable information. New templates should not be introduced merely to make activity visible at the centre.

Proportionality depends on risk, scale and maturity. A large national programme or high-stakes claim requires stronger evidence than a small exploratory activity. An institution with weak implementation may require more frequent follow-up, but the purpose and exit condition should be clear. Burden should not fall most heavily on institutions with the least administrative capacity unless additional support is provided.

Sampling can provide national evidence without universal collection. A representative study of learner work or institutional practice may answer a policy question more accurately than a census of self-reported compliance. Census data remain appropriate for basic administrative measures such as enrolment or formally assigned responsibility, subject to data quality.

Data should be collected once and used responsibly for multiple legitimate purposes where definitions align. Reuse should not change the meaning of a measure or disclose information beyond the authority under which it was collected.

Educational integrity requires accuracy, fair representation of evidence, appropriate treatment of uncertainty and respect for learners. Materials should identify sources and distinguish observed conditions, scientific explanation, projections and policy arguments. Where evidence is contested, the nature of the disagreement should be represented in proportion to its evidentiary basis rather than as an artificial balance between unequal claims.

Curriculum content should be reviewed for age appropriateness and unnecessary distress. Serious environmental and social questions can be taught without placing responsibility for distant or systemic harms on individual children. Learning should preserve agency while avoiding promises that simple personal action will resolve complex problems.

Political and commercial materials require scrutiny. Public bodies, enterprises and advocacy groups may possess useful expertise, but their interests and claims should be disclosed. Institutions should retain control of learner access, instructional purpose and assessment. Materials should not advertise products or collect learner data without proper authority and protection.

Teachers should be supported to manage disagreement. The standard is not neutrality between fact and falsehood; it is disciplined treatment of evidence and legitimate policy alternatives. Learners should be able to ask questions and reach reasoned conclusions without penalty for declining to express a prescribed personal belief.

Participation is both an educational method and a governance consideration. Learners can contribute to decisions about bounded institutional matters, undertake enquiry and present evidence. Communities can inform curriculum relevance, local knowledge and implementation conditions. Participation should have a stated purpose, transparent boundaries and feedback.

Representation requires care. A small council or project group should not be treated as the voice of all learners. Authorities should use accessible methods, include groups whose participation is often limited and report who was consulted. Participation costs, language and meeting arrangements can determine whose views are heard.

Consultation should occur early enough to influence design. A meeting held after decisions are settled may communicate policy but does not constitute participation in its formation. Authorities should publish the matters open to change and explain the outcome, including reasons where proposals are not adopted.

Community partnership should not transfer public financing or service duties to households. Contributions may be voluntary and valuable, but a minimum educational entitlement must not depend on a community's ability to raise funds or provide unpaid labour.

Monitoring has little value without a response to weak evidence. National policy should specify how findings lead to support, revision or enforcement according to the nature of the issue. A curriculum omission may require national amendment; weak local delivery may require professional support; unsafe practice may require immediate action; unequal resource access may require budget reallocation.

Corrective action should identify the finding, responsible authority, required change, resources, timetable, evidence of completion and follow-up. Recommendations directed generally to “stakeholders” do not establish accountability. Where several bodies share responsibility, one should coordinate and report.

Sanctions are not the first or only response. Institutions cannot correct structural shortages that public authorities have not addressed. Review should distinguish unwillingness, misunderstanding, lack of capability and lack of resources. The response should be proportionate and should protect learners from continuing harm.

System learning requires aggregation of recurring findings. If many institutions misunderstand the same requirement, the problem may lie in national design or guidance. Public reporting should therefore analyse patterns and record changes to the policy itself.

National reports should adopt explicit decision rules. A claim that a policy is **adopted** requires an authorised instrument with date and scope. A claim that it is **operational** requires assigned responsibility, resources and an effective date. A claim of **implementation** requires evidence from institutions or learners within the defined coverage. A claim of **outcome improvement** requires comparable outcome evidence and a stated basis for attributing change.

No essential dimension should be offset by strength elsewhere. Extensive curriculum documentation does not compensate for absence of teacher capacity; strong results in advantaged schools do not compensate for exclusion; institutional operations do not compensate for absence of learning. A profile preserves these distinctions.

Uncertainty should affect the strength of the conclusion. Where coverage is partial, the claim should be partial. Where definitions changed, trend language should be avoided or qualified. Where evidence conflicts, the report should present the conflict and investigate it.

The decision rules should be published before results. This reduces the incentive to select thresholds after seeing the data. Revisions may be necessary as the field develops, but their effect on comparability should be stated.

The EFA monitoring framework available at the opening of the Decade rejects a narrow equation of quality with either expenditure or test performance. It begins with learner characteristics and context, examines enabling inputs and teaching and learning, and considers outcomes in relation to educational purpose. This structure is directly relevant to education for sustainable development because the new commitment could otherwise be represented by whichever single measure is easiest to collect.[REF-08]

Inputs matter. Teachers, learning time, materials and facilities constrain what institutions can provide. They are not, however, self-interpreting. A textbook can be available but inaccurate or unused; a training course can be completed without changed practice; scheduled time can be lost or devoted to activity without conceptual progression. Monitoring needs to examine the route from resource to educational opportunity.

Outcomes matter for the same reason. A system cannot establish quality only by showing that it intended to provide learning. Yet outcome measures must correspond to purpose and opportunity. A reading or mathematics assessment provides essential evidence about foundational capability; it does not measure every civic, ethical or practical aim. A new ESD task may examine applied reasoning; it does not replace evidence about literacy, scientific knowledge or institutional conditions.

The framework therefore supports a chain of warranted conclusions. Policy and resources can establish readiness. Curriculum and practice evidence can establish opportunity. Learner work can establish specified capability. Distributional evidence can establish who benefited. Wider social change requires additional evidence and should not be attributed automatically to education.

The amount of time formally assigned to a curriculum area differs from time in which learners are engaged with relevant material. School closure, absence, late arrival, administrative interruption and weak organisation can reduce effective time. Within a lesson, unclear tasks or unavailable materials can produce further loss. National policy should distinguish scheduled, delivered and engaged time where evidence permits.

ESD implementation creates a further risk: time may be recorded because a project or event took place, while the intended knowledge and reasoning receive little attention. Evaluation should sample learner work and instructional sequences, not only calendars. An activity extending across several days should identify which curriculum expectations it advances and whether the time is proportionate to the result.

Integration can make productive use of existing time, but it should not be used to claim that no time is required. Teachers need time to prepare connections and learners need time to acquire and apply knowledge. Curriculum authorities should estimate the instructional demand and identify content that is consolidated, replaced or reorganised.

Time is also an equity measure. Institutions serving disadvantaged populations may experience more interruption, teacher absence or multi-grade complexity. The formal national entitlement may therefore produce less actual opportunity. Review should examine patterns rather than assume that a common timetable creates equality.

The quality evidence places teachers at the centre of improvement and records the scale of recruitment and preparation challenges associated with universal primary education. The policy implication for the Decade is not simply that more training is required. Teacher knowledge, pedagogical method, professional conditions and system support have to be considered together.[REF-08]

Structured teaching offers a useful corrective to the assumption that participatory education is unstructured. Direct explanation, guided practice and independent application can be combined in a purposeful sequence. The teacher establishes knowledge and criteria, supports learners to use them, and gradually transfers responsibility. Enquiry is strongest when learners know what question they are examining and possess the conceptual resources to interpret evidence.

Professional judgement is exercised within public expectations. Teachers should be able to select examples, respond to learner questions and adapt methods. They should not be required to determine the national entitlement individually. Clear curriculum and assessment guidance reduce uncertainty and allow professional effort to focus on teaching.

Preparation should include examination of misconceptions and contested evidence. A teacher asked to connect environmental, economic and social questions may encounter material outside the original specialism. Access to reviewed resources and subject networks is necessary. Collaboration should not be assumed to occur without time and leadership.

Learning materials shape the content that teachers can make available and the evidence learners encounter. Quality review should examine availability, accuracy, curriculum alignment, language, readability, accessibility and usability. Counting titles or copies addresses only part of the question.[REF-08]

Materials for education for sustainable development are especially vulnerable to rapid obsolescence, advocacy bias and transfer between contexts. Statistical statements require dates and sources. Case studies should identify their location and limits. Scientific explanations should reflect the evidence available, while policy alternatives should be presented as choices involving authority, cost and distribution.

Distribution systems can create unequal access even where national procurement is adequate. Delivery dates, storage, replacement and the ratio of usable copies to learners should be monitored. Digital distribution may reduce some costs but creates dependence on electricity, equipment, connectivity and technical support. At the present stage, national policy should not presume universal digital access.

Local production and translation can improve relevance. Quality controls remain necessary. Translation should preserve technical meaning and should be reviewed by speakers familiar with educational use. Community participation can identify inappropriate examples or omissions; final responsibility for accuracy and learner protection remains with the authorised provider or public body.

Learners cannot benefit fully from relevant content if the language of instruction prevents comprehension and participation. UNESCO's 2003 position paper connects language policy with educational quality, equality and cultural identity, and supports instruction that builds on the learner's language while enabling access to wider communication.[REF-13]

For ESD, language carries knowledge about place, livelihood, relationships and cultural practice. Excluding local languages can exclude knowledge and restrict community participation. Confining learners to a local language without a planned route to national or wider languages can also limit access to further education, scientific information and public life. National policy should address the sequence and resources required rather than present the choice as absolute.

Materials, teacher deployment and assessment must align with the language policy. A curriculum may authorise local-language learning while examinations or textbooks operate only in another language. This contradiction affects both opportunity and the validity of outcome measures. Performance may reflect language access rather than the capability the task is intended to assess.

Consultation should include language communities and teachers. Terminology may need development, but borrowing or translation should not be treated as evidence that a language is unsuitable for complex learning. Technical precision, intelligibility and progression can be developed through planned curriculum work.

Clean water, sanitation, safe structures and access for learners with disabilities are basic quality conditions. They also have direct relevance to sustainable-development learning. A school that discusses public health or resource responsibility while lacking minimum facilities exposes a structural contradiction. The response should be public investment and accountability, not a judgement that the school or learners failed to demonstrate commitment.[REF-08]

Facilities can support observation and practical work, but investment should follow educational need. Laboratories, gardens, workshops or monitoring equipment require safety, maintenance, teacher capability and equitable scheduling. A facility that is locked, unsafe or accessible only to selected learners does not establish opportunity.

Institutional operational data may provide learning material. Water or energy use can be examined through age-appropriate records where collection is lawful and safe. Learners should not be expected to audit complex systems without support, and sensitive security or personal data should remain protected.

Accessibility should be built into new materials and facilities. Retrofitting after implementation is often more costly and may delay participation. Consultation with affected learners and specialists can identify barriers not visible in design documents.

School leaders influence the organisation of time, professional collaboration, resource use and relationships with communities. The available quality evidence supports attention to leadership while warning against autonomy without resources and defined responsibility.[REF-08]

National policy should identify decisions that institutions may make and those retained centrally or regionally. Leaders can adapt implementation, support teachers and use evidence for improvement. They cannot correct national curriculum contradictions, teacher shortages or inadequate capital finance alone. Accountability should follow authority.

Leadership development should include curriculum interpretation, evidence use, participation, safeguarding and improvement planning. It should not create the expectation that leaders become specialists in every sustainability domain. Access to advice and networks remains necessary.

Institutional autonomy should be accompanied by transparency and equity safeguards. Well-resourced schools may use autonomy more readily, while weaker institutions require support. National monitoring should examine whether flexibility produces richer opportunity for some learners and a reduced entitlement for others.

PISA 2003 demonstrates that internationally comparable assessment can examine applied literacy and the distribution of performance, and can relate outcomes to learner and school characteristics. It also illustrates why scope must remain explicit: the study concerns 15-year-olds in participating countries and specified domains. It is not a universal account of education systems or of sustainable-development learning.[REF-11]

National authorities may learn from its attention to task design, sampling, background information and equity. They should not attach an ESD label to existing results merely because tasks involve real-world contexts. Construct validity depends on what was designed and measured.

Cross-national comparison of ESD policy may initially focus on definitions, curriculum location, teacher arrangements and monitoring methods. Even these require attention to differences in governance and terminology. Quantitative ranking should wait for evidence that measures have comparable meaning and sufficient coverage.

The absence of a global comparison should not prevent national baseline work. It should encourage publication of metadata and methods so that later cooperation begins from transparent evidence rather than retrospective reconstruction.

The policy context in sub-Saharan Africa is marked by the simultaneous demands of expanding access, improving completion, recruiting and preparing teachers, and strengthening basic learning conditions. Regional diversity is substantial; no single implementation description applies to every country. The common planning issue is that an expanded educational purpose must be pursued through systems in which institutional capacity and public finance are often under severe pressure.[REF-08]

National policy should connect ESD with priorities already present in sector plans. Water, health, agriculture, livelihoods, energy and local resource management can make foundational learning more relevant, but relevance should not replace structured development of literacy, numeracy and scientific understanding. Curriculum integration should identify a manageable sequence and provide materials in languages teachers and learners can use.

Teacher supply is likely to determine scale. New specialist staffing would be unrealistic in many settings and could divert personnel from basic provision. Initial and in-service education should therefore help existing teachers use locally relevant content within established subjects. Regional teacher-resource centres, radio and print may support reach where connectivity is limited, subject to evidence of use and follow-up.

Rural and remote institutions require explicit provision. Distribution of textbooks, teacher support and assessment materials can lag behind national adoption. Reporting should identify actual receipt and use by location. Demonstration activity in capitals should not be presented as national implementation.

Community participation may provide knowledge and support, but households should not carry costs that deepen exclusion. National and local authorities remain responsible for minimum conditions, educator preparation and learner protection.

Large populations, linguistic diversity, rapid expansion and pronounced inequalities shape implementation in South and West Asia. National averages can conceal major differences among regions, social groups and education pathways. Policy should establish a common entitlement while allowing subnational and linguistic adaptation.

Curriculum load and examination pressure require attention. Sustainable-development content may already appear across science, geography, social studies and value education. Mapping should determine whether the treatment is progressive and whether examinations encourage application. Adding a separate textbook without revising incentives may increase workload with limited effect.

Language policy affects access to local knowledge and wider information. Materials and teacher preparation should support the authorised language sequence. Assessment results should not be interpreted without considering whether learners encountered the content in a language they understood.[REF-13]

Urban growth, air and water conditions, agriculture, disaster exposure and employment transitions may provide relevant contexts. Selection should reflect national and local evidence and avoid presenting affected communities only through deficit. Learners need opportunities to examine public authority and structural conditions as well as household action.

Systems in East Asia and the Pacific vary from highly resourced examination-oriented systems to low-income and geographically dispersed provision. The regional consideration is therefore differentiation, not a common delivery model. National authorities should examine the relationship between strong subject performance, curriculum breadth, assessment pressure and opportunity for enquiry.

In systems with established comparative assessment and administrative data, monitoring can build on existing capacity. New constructs should still undergo validity work. Existing science or civic results should not be relabelled as ESD outcomes without evidence that the intended capability was measured.

Pacific island systems face scale, distance and exposure to coastal, water and weather-related risks. Regional cooperation may reduce the cost of curriculum expertise, materials and teacher development. Shared resources should allow local language, culture and national authority to shape final use.

Economic and technological change may create pressure for rapid curricular response. Quality policy should distinguish durable knowledge and capability from short-lived campaigns. Employer participation in vocational learning should be balanced by public-interest and learner-protection requirements.

Education systems in the Arab States differ in resources, governance, participation and exposure to conflict or water scarcity. National design should reflect these differences and should not assume that a common language eliminates variation in curriculum access or institutional capacity.

Water, urban development, energy, health and employment may provide substantive contexts. Materials should connect scientific and technical evidence with economic and social consequences, while maintaining space for reasoned policy difference. Translation of international material requires review for terminology, relevance and educational level.

Gender participation and progression should remain visible in reporting. National policy should examine whether curriculum pathways, facilities, transport, teacher availability or social conditions produce different opportunity. ESD projects should not be used to claim inclusion where broader participation barriers remain.

Conflict-affected settings require a priority on continuity and safety. Regional or national materials designed for stable institutions may need substantial adaptation. Public reporting should distinguish emergency provision from ordinary curriculum implementation.

Latin America and the Caribbean combine high urbanisation in many countries with rural, indigenous and environmentally diverse settings. National quality policy should address inequality within systems and the relationship between national curriculum, local knowledge and language.

Decentralisation may permit contextual adaptation but can also produce unequal capacity. Central authorities should define minimum entitlement, evidence and financing arrangements. Local authorities and institutions can select relevant examples and partnerships where responsibility and safeguards are clear.

Indigenous knowledge and cultural identity require participation by the communities concerned. Curriculum should avoid extracting isolated practices or presenting culture as static. Learners should gain access both to community knowledge and to wider scientific, historical and civic understanding.

Caribbean and other small systems may benefit from regional cooperation in materials, examinations and teacher education. Shared arrangements should disclose which body holds authority and how national context is preserved. Small sample sizes require cautious public comparison.

Many systems in Europe and North America have extensive curriculum, data and professional infrastructure, but this does not remove the need for policy coherence. Strong subject provision can remain fragmented, and high-stakes selection can narrow learner experience. Socio-economic and geographic inequality can affect access to applied and participatory learning.[REF-11]

Curriculum review should identify whether sustainable-development questions are treated with sufficient disciplinary depth and whether learners can connect evidence across fields. Institutional projects should be evaluated for learning rather than visibility. Access to technology or external partnerships should not determine the minimum entitlement.

Public debate may be politically contested. Materials and assessment should maintain evidentiary standards, distinguish scientific findings from policy choices and allow reasoned disagreement. Learners should not be assessed on adherence to a preferred personal position.

These systems may also examine the international consequences of consumption, production and policy. Such study should use traceable evidence and avoid portraying other regions as undifferentiated recipients of harm or assistance.

Regional analysis should identify shared implementation conditions, sources of cooperation and distributional concerns. It should not replace country analysis. Administrative structures, languages, fiscal capacity and curriculum authority vary within every region.

Comparable reporting can begin with a small common core: policy status, scope, teacher support, curriculum entitlement, data coverage and stated limitations. Narrative profiles should explain context and national priorities. Quantitative comparisons should use common definitions and reference periods or remain clearly non-comparable.

Regional bodies and networks can support exchange, joint research and shared technical resources. Participation should not be represented as national implementation, and shared guidance should not be confused with domestic authority.

Chapter 36 of Agenda 21 remains the broadest foundation for the educational dimension of sustainable development. Its three programme areas—reorienting education, increasing public awareness and promoting training—establish a scope extending across formal education, non-formal learning, public communication and workforce capability. For national quality policy, its significance lies less in a detailed set of school requirements than in its conception of education as a cross-cutting means of implementation.[REF-01]

The chapter links basic education with wider environmental and development learning. This relationship prevents two misreadings. The first would treat education for sustainable development as an advanced curriculum available only after universal basic education has been achieved. The second would divert attention and resources from literacy, access and minimum learning conditions. The proper interpretation is that basic capability enables wider learning and that relevant content can strengthen, but not replace, the quality of basic provision.

Agenda 21 also connects formal and non-formal education. A national plan limited to school curriculum would therefore address only part of the mandate. Public awareness and training require separate purposes, audiences and quality controls. Public information may seek informed behaviour or participation; vocational training may seek safe and competent performance; school education has broader developmental and rights-based purposes. One reporting framework should not assume identical evidence across these forms.

The call to integrate environment and development concepts across educational programmes does not imply that every subject must carry the same content. Integration can be established by identifying appropriate disciplinary contributions and ensuring that learners can connect them. National review should look for coherent coverage and progression rather than the repeated presence of a common vocabulary.

The chapter's attention to scientific evidence and local context is especially significant. National policy should prevent locally relevant curriculum from becoming anecdotal or inaccurate, while ensuring that internationally produced material is not applied without contextual examination. The quality standard is a disciplined relationship between reliable knowledge and the circumstances in which learners use it.

Agenda 21 includes resource and capacity considerations. Its policy logic is incompatible with an implementation model that imposes duties without teacher preparation, materials or institutional support. Even where national governments choose different instruments and schedules, the means of implementation belong within the quality judgement.

The Dakar Framework for Action situates quality within Education for All. Its goals encompass early childhood, primary education, learning needs of young people and adults, literacy, gender equality and the quality of education. The quality goal refers to recognised and measurable outcomes, especially in literacy, numeracy and essential life skills. National ESD policy should be read through this existing commitment rather than developed as an unrelated programme.[REF-06]

The term “recognised and measurable” does not require every educational purpose to be converted into a national test. It requires public clarity about the outcomes sought and credible evidence concerning them. Some outcomes may be assessed through sample studies, moderated work, programme evaluation or institutional evidence. The method should be appropriate to the construct and use.

The Dakar framework reinforces the national planning responsibility. ESD measures should appear in the education-sector plan where they affect curriculum, teachers, finance or data. A separate strategy can provide focus, but it should state how it is authorised, funded and integrated with the sector plan. Otherwise, competing timetables and reporting demands may arise.

Gender equality has immediate relevance. The 2005 target concerns disparity in primary and secondary education, while the broader aim concerns equality in access, achievement and educational experience. ESD reporting should therefore examine participation and opportunity by sex and should avoid treating nominal access to a curriculum as evidence of equal benefit.

Literacy and essential life skills provide another connection. Learners require literacy to interpret environmental, health, economic and public information. Adult literacy programmes can use relevant content, but they should develop transferable reading, writing and numeracy rather than train participants to repeat a fixed message.

The Johannesburg Plan of Implementation renews educational commitments within a wider programme of poverty eradication and sustainable development. It recalls universal primary education and gender equality and calls for the integration of sustainable development into education systems at all levels. It also addresses national and local plans, research, information, skills and public awareness.[REF-02]

For education authorities, the Plan creates a coordination requirement. Development ministries and agencies may identify legitimate knowledge and capability needs, but curriculum decisions require educational analysis. A national coordination mechanism should translate public priorities into age-appropriate, evidence-based learning rather than transmit sector messages directly to institutions.

The Plan's connection with poverty eradication requires distributional review. Learners most affected by environmental degradation, unsafe work or insecure access to services may attend the least-resourced institutions. An implementation model dependent on technology, travel or voluntary household contributions could therefore direct the richest learning opportunities away from those whose circumstances are most relevant to the subject.

The Plan also supports a relationship between education and national sustainable-development planning. Education should not be described only as a tool for obtaining public acceptance of development decisions. It prepares people to understand, scrutinise and participate in those decisions. Quality policy should protect this educational and civic function.

Resolution 57/254 proclaims the Decade for 2005–2014 and designates UNESCO as lead agency. The resolution establishes international direction and coordination; it does not assign UNESCO the domestic powers of national education authorities. National documents should preserve this distinction.[REF-03]

The lead-agency role can support guidance, coordination and international reporting. Governments remain responsible for legislation, curriculum, teacher employment, finance and institutional oversight according to national arrangements. Claiming that a domestic measure is required by UNESCO when it is in fact a national choice would obscure accountability.

The proclamation also provides a defined period. A ten-year horizon permits phased implementation and evaluation. It does not justify postponement without baseline action, nor does it require every outcome to be achieved immediately. National plans should identify what evidence is expected at early, intermediate and later stages.

The period supports continuity across political and budget cycles. Authorities should place responsibilities in ordinary institutions and plans rather than rely on a launch committee that may not persist. Review dates and public reporting can maintain attention without creating a permanent parallel bureaucracy.

The later resolutions concern preparation and implementation of the Decade. Resolution 59/237 is especially relevant at the date of this report because it confirms the start on 1 January 2005, reaffirms the educational commitments connected with the Decade and addresses completion and consideration of the international implementation scheme. It encourages governments to consider measures within education systems, strategies and national development plans.[REF-04] [REF-05]

The language of national consideration leaves room for different legal and administrative forms. It does not leave room for ambiguity about whether an authority has considered the issue. A government should be able to record the responsible process, evidence reviewed, decision reached and relationship to existing plans. The outcome may be amendment of current policy rather than creation of a new instrument.

The resolution's treatment of the implementation scheme is a control on status. At the cut-off date, the October 2004 scheme is a draft awaiting the identified institutional process. National authorities may draw on it for planning, but should not cite it as a final adopted standard. This distinction affects references, legal claims and public communication.

The resolution also recognises existing educational initiatives. National implementation should avoid claiming that the Decade supersedes Education for All or other established commitments. It provides an opportunity to connect them more coherently.

The draft International Implementation Scheme available at the commencement of the Decade describes an ambitious partnership approach and a set of strategies including advocacy, consultation, partnership, capacity building, research, innovation, information and monitoring. Its value for national policy lies in the breadth of implementation functions it identifies. Its limitation lies in its current status and level of generality.[REF-07]

National authorities should distinguish a strategy from an evidence standard. Advocacy can build understanding, but awareness of a policy is not proof of educational implementation. Networking can exchange practice, but participation in a network does not establish quality. Capacity-building activity must be connected to demonstrated professional capability. Monitoring requires defined indicators and baselines.

The partnership approach can mobilise expertise, yet public responsibility should remain clear. Civil society, communities, research bodies and enterprises may contribute. National quality policy should establish selection, disclosure, curriculum control, learner protection and continuity. A partnership should not determine entitlement according to local fundraising capacity or commercial access.

The draft scheme's broad conception of sustainability includes social, economic, environmental and cultural considerations. Curriculum frameworks should preserve this breadth while setting priorities appropriate to age and context. Breadth does not justify indefinite scope; progression and subject allocation remain necessary.

The final scheme may alter wording, emphasis or arrangements. A national plan based on the draft should include a review point after the relevant international decision. Historical records should preserve which version informed the original decision.

Articles 28 and 29 of the Convention provide a rights foundation. The right to education concerns access and progression, while the aims of education include development of the child's abilities, respect for human rights and cultural identity, preparation for responsible life, and respect for the natural environment. The Committee's General Comment No. 1 treats these aims as relevant to content, processes, pedagogy and the educational environment.[REF-09] [REF-10]

This record supports a broad quality concept. It also sets limits on implementation. Education should develop the child's capacity and dignity; it should not compel personal declarations or use learners as instruments for institutional publicity. Participation should be age-appropriate, safe and meaningful. Discipline and institutional practice should be consistent with the values taught.

Respect for cultural identity and the natural environment should be considered together. Curriculum should not treat culture as an obstacle to scientific learning or use environmental claims to erase legitimate cultural difference. It should enable examination of practices and consequences through reliable evidence and respect for persons.

The rights framework strengthens the equity test. A national curriculum that exists formally but is inaccessible because of language, disability, cost or exclusion does not provide an equal educational opportunity. Monitoring should address barriers and the distribution of conditions, not only average outcomes.

The 2005 EFA Global Monitoring Report provides analysis of quality, learning conditions and international progress. It is not a treaty, resolution or national regulation. National policy should use its evidence and conceptual framework without misrepresenting it as a binding legal standard.[REF-08]

Its quality framework is useful because it connects learner characteristics, context, inputs, teaching and learning, and outcomes. This supports the profile approach proposed in this report. It also helps prevent policy from equating a curriculum statement with quality or a test result with the full educational purpose.

The report's evidence on teachers, learning time, materials, facilities, pedagogy and leadership identifies implementation conditions that should be examined in ESD planning. Those conditions vary and should not be converted into universal thresholds without national evidence. A pupil-teacher ratio, for example, describes one condition but does not by itself establish the quality of teaching.

The EFA analysis also places equity at the centre. National ESD monitoring should be linked with access, completion and foundational learning data. A programme may be educationally strong for participants and still fail a public-interest test if its design systematically excludes others.

The first results from PISA 2003 provide comparative information about the capacity of 15-year-olds in participating countries to apply reading, mathematics and science knowledge, together with background information and distributional analysis. They demonstrate the policy value of examining application and equity.[REF-11]

Their use in this report is methodological. PISA does not measure the national implementation of the Decade, which began after its data collection. It does not cover every country, age group or ESD purpose. No result should be interpreted as an ESD ranking.

The study illustrates controls relevant to future measurement: a defined population, explicit domains, standardised tasks, sampling, background variables and published methods. National ESD assessment should adopt comparable discipline while developing constructs suited to national curriculum and purpose.

The evidence also shows why a single average is insufficient. Distribution by social background and other characteristics matters to public policy. Any future ESD outcome measure should examine opportunity and performance across groups, subject to data quality and protection.

The sources used in this report perform different functions. Treaties establish obligations for States Parties. General Assembly resolutions record collective decisions and expectations. Conference plans and frameworks establish agreed policy direction. Draft instruments inform development but retain provisional status. Monitoring reports and comparative studies provide analysis and evidence. National laws and policies provide domestic authority.

These functions should not be merged. A finding from a monitoring report does not become a legal duty because it appears in an official publication. A broad international commitment does not supply the detailed evidence needed for a national curriculum decision. A national choice should not be attributed to an international body merely to strengthen its appearance of authority.

Source: ICEQC interpretation of the status and function of the evidence base used in this report.[REF-01] [REF-02] [REF-05] [REF-07] [REF-08] [REF-09] [REF-11]

An evidence register should record source type and role. Reviewers should test each material claim against the source that supports it and should use the strongest available authority for legal status and the strongest available evidence for empirical claims. One source need not perform both functions.

Assume that a national authority intends to revise its curriculum framework during 2005 and wishes to incorporate the Decade. The existing curriculum includes environmental topics in science and geography, community participation in civic education and resource questions in vocational programmes. Coverage is uneven across stages, and no common progression or assessment expectation exists.

The first decision is whether the policy problem is absence of content or absence of coherence. A documentary mapping may show substantial existing content but limited connection, repetition of simple awareness messages and no clear entitlement for learners outside selected subjects. In that case, adding a new compulsory theme would not address the principal weakness.

The authority should define a small set of capabilities and map them to phases and subjects. Subject panels should identify essential knowledge, while a central group checks progression, pathway coverage and overlap. The framework should distinguish requirements from examples. Institutions need to know which expectations are mandatory and where professional adaptation is permitted.

Consultation should test intelligibility and feasibility. Teachers can identify unrealistic time or knowledge demands; learners can identify inaccessible language and whether the intended opportunity is visible; communities and sector experts can identify relevance and accuracy. Consultation does not determine scientific validity by vote. Evidence claims remain subject to specialist review.

Assessment review should occur at the same time. Where national examinations govern progression, specifications should indicate how application and reasoning can be recognised within subjects. Any new task type requires piloting before it affects high-stakes decisions. Classroom guidance should provide examples without prescribing a single lesson sequence.

The implementation record should state the baseline, documents revised, teacher and material implications, effective dates and populations covered. The first claim may properly be that the curriculum entitlement has been adopted. Evidence of taught opportunity and learner capability belongs to later review.

Assume that curriculum policy is already adopted but teachers report uncertainty about content and method. The government proposes a one-day orientation for all teachers. Such an event may communicate the mandate, but it is unlikely to establish the capability required for planning, teaching and assessment.

The authority should disaggregate the capability. All teachers may require an account of national purpose and curriculum responsibilities. Primary teachers may require integrated planning and age-appropriate enquiry. Secondary teachers require subject-specific content and coordination. Vocational educators require occupational application. Leaders require planning, support and evidence use. Teacher educators and advisers require deeper preparation to sustain the programme.

A staged model could combine an initial briefing, subject or phase modules, supported planning, use in practice and review of learner work. Completion should require engagement with the applied component, not only attendance. The programme should provide printed or otherwise accessible materials and should not assume connectivity.

Trainer preparation is a quality control. A cascade model can extend reach but may simplify content and weaken consistency at each transfer. Core materials, trainer criteria, observation of a sample of delivery and access to advice can reduce this risk. Regional adaptation should be recorded rather than treated as deviation automatically.

Coverage reporting should use teachers within scope as the denominator and show phase, subject, location and other relevant characteristics. Evaluation should examine knowledge, planning and practice on a sample basis. Learner outcomes should not be attributed to the programme before sufficient exposure and comparison.

Workload and replacement require financing. If teachers attend during instruction, learner time may be lost; if they attend outside contracted time, participation and morale may suffer. The plan should identify the arrangement and cost rather than treating teacher time as free.

Assume that the ministry asks every school to conduct a self-review of its contribution to education for sustainable development. Without a defined scope, schools may compile activity lists, photographs and policy statements that are difficult to compare and weakly connected to learning.

The review should begin with learner entitlement. Schools identify where the national curriculum is taught, which learners receive it and what work shows the intended progression. They then examine enabling conditions: teacher preparation, materials, time, participation, safety and community access. Operational practices are considered where they contribute to learning or contradict stated aims.

Evidence should be proportionate. A school may use curriculum plans, a sample of learner work, assessment tasks, professional-learning records, learner discussion and relevant operational records. It should not be required to create documents that duplicate existing evidence. The self-review should distinguish strengths, constraints within school control and structural needs requiring action by another authority.

The outcome is an improvement plan with a limited number of actions, owners, resources and review dates. A school unable to provide safe water should record the effect and refer the matter to the responsible authority; it should not be judged as if the capital failure were a discretionary school choice.

External review may sample the quality of self-evaluation and corroborate evidence. The purpose is not to reward persuasive presentation. Reviewers should examine whether conclusions follow from evidence and whether learners with different backgrounds receive the opportunity.

National aggregation of self-review categories can identify recurring constraints, but narrative judgements should not be converted mechanically into a league table. Differences may reflect candour, evidence capacity or context as well as performance.

Assume that the authority wishes to measure whether lower-secondary learners can reason about a resource-allocation question. The intended use is programme evaluation, not individual certification. The task will be administered to a sample after the revised curriculum has been taught.

The construct should be stated before task development. Relevant components may include interpretation of a data table, identification of affected groups, comparison of alternatives, recognition of uncertainty and justification of a feasible recommendation. The task should supply essential information and should not assume personal experience or a preferred political conclusion.

Content specialists verify accuracy; curriculum specialists verify opportunity; assessment specialists examine task demand and scoring; language and inclusion reviewers identify barriers. Learner trials investigate comprehension, time and unintended cues. Scoring examples should include different defensible conclusions supported by evidence.

The sample design should represent the population to which conclusions will apply. Exclusions, non-response and school participation should be reported. Where curriculum implementation varies, contextual data can identify whether learners had an opportunity to learn. The assessment should not be used to sanction schools during the pilot.

Analysis should report distributions and relevant group differences with uncertainty. A national mean cannot show whether performance is broadly shared. Item-level evidence may reveal which aspects of reasoning are difficult, but post hoc interpretations should be tested against learner responses.

The pilot report should identify whether the measure is sufficiently valid for continued use, which revisions are needed and what cannot be concluded. An unsuccessful instrument should be withdrawn or redesigned rather than retained to preserve a public commitment.

Assume that a government has a fixed transition budget for curriculum materials and teacher support. Equal per-school allocation is administratively simple, but schools differ in size, distance, language needs, existing resources and teacher preparation.

The authority should first cost the minimum implementation package. Fixed costs, such as one reference set or a local planning session, differ from learner-related costs. Remote delivery, translation and accessibility may add legitimate cost. Allocation should combine a base amount with factors reflecting enrolment and need.

Need classifications require transparent data and regular review. A deprivation measure may be relevant, but it should not be assumed to capture distance, disability access or language cost. Authorities should test how the formula distributes funds and should publish material assumptions.

Allocation is not expenditure. Financial monitoring should record receipt, timing, permitted use and actual spending. Institutions may be unable to use funds if procurement rules, supply or staffing are inadequate. Support should address the constraint rather than interpret underspending automatically as lack of commitment.

Evaluation should examine whether the allocation enabled the intended resource and whether learner opportunity improved. A cost-effectiveness comparison must preserve quality and equity requirements. A cheaper model that excludes remote schools is not more efficient for a universal policy.

Assume that national government defines broad educational aims while regions control curriculum and teacher support. The Decade may produce a national declaration without a clear implementation mechanism, or multiple regional approaches without a common basis for public reporting.

The national and regional levels should agree which elements form the common entitlement and which are adaptable. Common elements may include purpose, minimum capability, rights and safeguards, core metadata and reporting dates. Regions may determine content examples, subject allocation, training delivery and additional indicators.

Authority and funding should be documented. If national reporting is required, regions need resources and agreed definitions. National bodies should not collect comparable figures from measures that regions define differently. Where differences remain, the report should use profiles rather than a combined statistic.

Intergovernmental coordination should resolve specified decisions and maintain a record. It should not become a forum in which responsibility for delay is exchanged. Each action should have an owner, authority and date.

Regional innovation can inform wider policy. Evaluation should describe context and avoid requiring uniform adoption. A method effective in a densely populated region may not transfer to remote or linguistically different settings. Shared learning requires evidence about conditions, not only a description of the activity.

Assume that a public body commissions community workshops and media material on water use and local health. The programme contributes to the public-awareness aspect of Agenda 21 and the Decade, but it should not be reported as formal curriculum implementation.[REF-01]

The commission should define audience, purpose, content, language, delivery and intended learning. Technical statements should be reviewed against current official evidence. Material should distinguish public requirements, recommended practices and matters of individual choice.

Access may require radio, print, community meetings or workplace delivery rather than a single format. Attendance or estimated audience reach provides activity evidence. Evaluation should examine comprehension and usable capability on a sample basis. Later changes in water use or health have multiple causes and require a stronger design before attribution.

Facilitators need preparation to handle questions and recognise the limits of their role. They should not provide health or technical advice beyond the approved material. Referral routes should be available for matters requiring public services or specialist assistance.

Participation should not be coercive. Access to public services or benefits should not depend on expressing agreement with programme messages. Feedback can identify unclear information, local constraints and matters requiring action by authorities.

The seven cases concern different instruments and evidence. Curriculum revision establishes entitlement; professional learning develops capacity; self-review examines institutional implementation; assessment studies learner capability; finance enables delivery; decentralised coordination allocates authority; adult provision serves a different audience and purpose. Treating them as interchangeable “ESD activities” would prevent meaningful accountability.

Across the cases, the same controls recur for substantive reasons: defined purpose, lawful authority, evidence appropriate to the claim, resources, distributional review, transparent status and corrective action. The practical form changes with the decision. A national framework should preserve the controls without forcing every initiative into an identical administrative template.

Source: ICEQC applied analysis for this report.

Sustainable development concerns many domains. A curriculum that attempts to include every relevant topic will become unmanageable and may produce shallow treatment. A curriculum that selects too narrowly may omit the relationships that give the field its educational significance. National quality policy must manage breadth through progression and representative depth.

The minimum entitlement should specify capabilities and a limited body of essential knowledge. Examples can vary by context. Across stages, learners should encounter more than one type of question and should examine relationships among social, economic and environmental factors. They need not study every sector.

Depth requires sufficient subject knowledge, instructional time and assessment. An interdisciplinary project should be selected because the question requires several forms of knowledge, not because breadth is itself evidence of quality. Curriculum review should identify whether connections are analytically necessary and whether learners possess the underlying concepts.

Authorities should disclose prioritisation. Removing or consolidating content is a policy decision and should be based on educational purpose, not concealed through language suggesting that all existing and new expectations can be accommodated without cost.

Local context makes learning intelligible and allows learners to examine direct evidence. A common national entitlement protects learners from receiving a narrower education because of geography, institutional resources or local preference. These aims are complementary if policy distinguishes capability from example.

National frameworks can specify knowledge and reasoning that every learner should develop. Regions and institutions can select contexts suited to language, environment, livelihoods and available evidence. Review should determine whether adaptations meet the same intellectual demand and provide access to wider national and global relationships.

Local knowledge should enter through respectful participation and reliable representation. A community is not internally uniform, and one participant should not be treated as its sole voice. Curriculum should identify the source and context of knowledge and permit examination alongside other evidence.

Common requirements should not erase cultural identity. Rights instruments support both broad educational aims and respect for language and culture. National policy should explain how the entitlement is realised in diverse settings rather than assume that uniform materials are neutral.[REF-09] [REF-10] [REF-13]

Participation enables learners and communities to contribute knowledge, identify consequences and practise public reasoning. Expertise supports accuracy, safety and technical judgement. Policy fails if it treats either as sufficient on its own.

Experts should state the scope and limits of their evidence. They should not determine value judgements outside their competence merely because they provide technical advice. Participants should have meaningful opportunity to question assumptions and describe local conditions. Their experience does not remove the need to test general claims.

Decision processes should identify who decides. A learner consultation can inform school policy; it does not transfer the head teacher's safeguarding responsibility. Community advice can shape curriculum examples; national authorities retain responsibility for entitlement and accuracy. Technical panels can review evidence; elected or authorised bodies make public policy according to domestic arrangements.

Records should show how input affected the decision. Participation loses credibility when contributions are collected without response. Conversely, authorities should explain why a proposal was not adopted where evidence, rights, cost or feasibility provided a reason.

Environmental and social pressures can create justified urgency. Education systems nonetheless require careful curriculum, teacher and assessment development. Rapid action that produces inaccurate materials, unsupported duties or invalid measurement can weaken the policy and public confidence.

Authorities should distinguish measures that can begin immediately from those requiring development. Immediate measures may include assigning responsibility, mapping existing provision, protecting accurate public information and commencing teacher consultation. Curriculum progression, assessment instruments and national outcome claims require more time.

Piloting is not delay when it tests a material uncertainty. It should have a timetable and decision point. Endless experimentation without a decision can become avoidance, while national scale before feasibility evidence transfers risk to learners and teachers.

Public reporting should explain the sequence. A restrained statement of what has and has not yet been established is more credible than a claim of completed integration at the beginning of the Decade.

Consistency protects entitlement and comparability. Autonomy enables institutions to respond to learners, context and professional judgement. The policy question is which decisions require common rules and which benefit from local determination.

Common rules are strongest where rights, minimum curriculum, safety, assessment validity, data definitions and public finance are concerned. Local discretion is valuable in examples, teaching sequences, partnerships and improvement methods, within those boundaries.

Autonomy requires capacity. An instruction to innovate may advantage institutions with experienced staff, flexible budgets and external partners. Authorities should provide support and monitor distribution. Flexibility should not permit a lower curriculum for learners in disadvantaged schools.

External review should examine outcomes and evidence rather than conformity with one preferred method. Where an institution chooses a different approach, it should be able to show how the national purpose is met and how risks are controlled.

Innovation can reveal new methods and mobilise professional energy. It often begins in institutions with stronger leadership, resources or networks. If policy scales by celebrating those institutions without analysing conditions, innovation may deepen inequality.

Pilot selection should include varied settings and should provide the support necessary for a fair test. Evaluation should record adaptations, additional resources and professional time. A model tested with exceptional staffing should not be costed as if ordinary conditions applied.

Scaling may require redesign. A simpler method supported nationally can produce wider educational value than a complex model sustained only through external expertise. Simplification should preserve the essential mechanism; evaluation should test that it does.

Intellectual and commercial ownership should be considered. Materials developed with public funds should be available on terms that permit equitable use and adaptation, subject to applicable law. Institutions should not become dependent on a provider where continuity or affordability is uncertain.

Accountability requires public evidence, but punitive use of immature measures can discourage honest reporting and experimentation. The first phase of the Decade should establish a learning-oriented evidence system while maintaining firm action on safety, rights and deliberate misrepresentation.

Indicators should be used according to validity. Curriculum mapping can support national review; it should not rank teachers. Sample assessment can identify system patterns; it should not determine individual consequences. Institutional self-review can guide improvement; external verification can examine whether conclusions are evidence-based.

Professional candour should be protected. Teachers and leaders should be able to report that a method was impracticable or a resource arrived late without being treated automatically as responsible for the failure. Authorities should distinguish evidence about system design from evidence about institutional conduct.

Public reporting should include corrective action by the central authority as well as expectations for institutions. Accountability is incomplete if only the delivery level is visible.

Comparable evidence permits national oversight and distributional analysis. Contextual validity ensures that the evidence has meaning in the setting examined. A common indicator can support both if its core definition is stable and contextual information accompanies it.

Some measures are suitable for common reporting: legal status, curriculum scope, teacher participation under a defined programme, resource receipt and selected learner outcomes from a common instrument. Others require profiles: local curriculum content, partnership arrangements, institutional improvement and barriers specific to language or geography.

Authorities should not force qualitative findings into numeric categories solely to aggregate them. Coding may be useful where definitions and reviewer reliability are established. Narrative evidence should remain available to explain mechanism and context.

Comparability over time requires version control. Changes in curriculum, sampling, task design or administrative boundaries should be documented. A new measure may be better while creating a break in trend; reporting should state both facts.

These tensions cannot be removed by a general call for balance. Decisions require criteria. National authorities should state the educational purpose, rights involved, evidence, affected population, feasibility, distributional effect and the consequence of delay or error. They should record alternatives considered and the reason for selection.

Source: ICEQC decision framework for this report.

The opening year should establish the evidentiary point of departure. A baseline is not a declaration that no relevant activity existed before the Decade. It records the condition of policy, curriculum, capacity, opportunity and available outcomes using defined methods and a stated reference period.

The baseline should incorporate existing data where definitions and coverage are adequate. New collection should address material gaps rather than attempt an immediate census of every possible activity. Curriculum documents, teacher requirements, national assessments, administrative data and a sampled institutional study can provide complementary evidence.

The report should identify which findings are national estimates, administrative counts, documentary judgements or illustrative cases. Later comparisons must use the same category or explain the change. The baseline should be preserved as a controlled publication even when methods improve.

An early review should examine whether responsibilities, resources and support reached the intended levels. It should focus on implementation processes and should not expect mature learner outcomes. Questions include whether guidance was understood, materials were usable, professional learning was completed and adaptations were recorded.

The review should include varied institutions and should examine burden and unintended effects. It should recommend changes with assigned authority. Public reporting should distinguish measures already corrected from those requiring longer action.

A mid-period evaluation should examine whether the national model produces the intended opportunity at scale and whether access is equitable. By this stage, learner evidence may be appropriate where curriculum exposure has been sufficient and measures have passed validity review.

The evaluation should revisit the theory of action. Some assumed relationships may not have operated. Teacher preparation may be adequate while assessment remains contradictory; curriculum coverage may improve while disadvantaged regions lack materials. Analysis should identify the point at which the chain weakens.

The mid-period report should inform revision, not merely record progress. Requirements, funding and indicators may need change. Any break in data series should be documented and overlapping measures retained temporarily where feasible.

The final assessment should distinguish what changed during the Decade from what can be attributed to specific policies. It should compare the baseline and later evidence, examine distribution, document institutional development and identify unresolved gaps.

Outcome claims should be proportionate to design. National trends can be described where measures are comparable. Causal claims require a stronger basis. Wider environmental, economic or social change should not be attributed to education without evidence connecting exposure, capability and subsequent effect.

The assessment should preserve negative and mixed findings. It should record the cost and sustainability of arrangements, including which depended on temporary support. The purpose is to inform the next policy period, not to close the record with a ceremonial statement.

A decade-long programme requires stable identifiers, definitions and archives. Authorities should maintain versions of policy documents, indicator specifications, instruments, datasets and reports. Revision records should show when a measure changed and which publications used it.

Staff turnover creates risk to institutional memory. Methods should be documented sufficiently for a new team to reproduce core processes. Data access and protection arrangements should be assigned to institutions rather than individuals.

Public accessibility should persist. Links and formats may change, but controlled copies and bibliographic records should remain. Historical reports should not be silently updated to reflect later knowledge.

Source: ICEQC monitoring architecture for this report. Dates should be determined nationally according to implementation sequence and evidence maturity.

The ministry or public authority responsible for education should hold ownership of the national educational response. Cross-government participation does not alter that responsibility. Ownership includes defining the educational purpose, approving curriculum and assessment arrangements within domestic authority, coordinating finance, commissioning evidence and reporting publicly.

The responsible minister or senior official should identify an accountable administrative unit. A broad committee without a permanent secretariat may support consultation but cannot substitute for an office able to manage decisions, records and follow-up. The unit should possess access to curriculum, teacher, finance and data functions; it need not duplicate them.

Terms of reference should specify decisions, reporting lines and duration. If the arrangement is established for the Decade, review points should determine whether its functions have been absorbed into ordinary administration. The objective is sustained responsibility, not preservation of a temporary structure.

Ministerial ownership also requires accuracy in public claims. Communications should be cleared against the evidence register and should distinguish national action from the role of international organisations. A public launch should not precede agreement on what institutions are expected to do.

Curriculum authorities should translate national purpose into entitlement, progression and subject responsibility. They should maintain a cross-curriculum map and resolve omissions or duplication. Subject panels remain responsible for disciplinary accuracy and appropriate sequencing.

Assessment authorities should examine the consequences of existing examinations and design any new measures according to stated use. They should participate early in curriculum development rather than receive a completed framework that cannot be assessed credibly. Assessment should sample the intended capability without defining the entire curriculum.

The two functions need a formal point of coordination. Joint records should identify the curriculum expectation, evidence available through classroom assessment, any national measure, and the timetable for change. Where assessment remains unchanged during a transition, guidance should explain the effect so that schools do not infer incompatible priorities.

Authorities responsible for initial teacher education should map programme requirements against the revised entitlement. The map should identify common professional knowledge, subject or phase specialisation, placement opportunities and assessment of applied competence. Approval or funding arrangements should be revised according to domestic authority.

Continuing professional learning should connect national, regional and institutional provision. A national framework can define purpose and core content; regional bodies can organise access and adaptation; institutions can support planning and review of work. Participation data should move upward without requiring repeated reporting by teachers.

Teacher representatives should be involved in workload, feasibility and professional standards. Consultation does not remove the public authority's duty to establish curriculum. It improves the evidence on implementation conditions and supports clarity about professional responsibility.

Subnational authorities may control curriculum detail, teacher deployment, institutional support, capital provision or data. The national operating model should list these functions and the finance attached to them. A responsibility matrix should be approved before implementation reporting begins.

Local support should use evidence to address barriers. Advisers can review curriculum planning, facilitate teacher collaboration and help institutions interpret findings. Their work should not become an additional inspection system with different criteria.

Where local authorities commission partners or materials, national safeguards concerning accuracy, disclosure, access and learner protection should apply. Local innovation should be documented sufficiently for evaluation and possible wider use.

Institutions should translate the entitlement into teaching, support and improvement within their authority. Their responsibilities include curriculum planning, assignment of staff and time, safe delivery, use of evidence, participation arrangements and referral of structural constraints.

Institutions should not be required to create a separate ESD plan if the relevant actions can be incorporated into curriculum and improvement planning. A parallel plan may fragment responsibility. The governing question is whether the intended opportunity and conditions are visible and reviewed.

Leaders should identify who coordinates interdisciplinary matters without assigning all responsibility to one enthusiastic teacher. Subject and phase leaders retain their roles. The coordinator facilitates coherence, evidence and links with institutional operations or community expertise.

The national statistical function should approve indicator definitions, denominators, classifications and quality controls. It should advise on whether administrative data can support the claim intended and should publish limitations. Research teams may develop new measures and investigate mechanisms; their exploratory work should remain distinguishable from official series.

Data collection should be planned across the Decade. Institutions should receive a consolidated schedule and clear definitions. Changes should be piloted and documented. The statistical body should retain the authority to withhold or qualify a figure that does not meet the required standard.

Research commissioning should state the question, method, independence, access to evidence and publication arrangements. Evaluation findings should not be suppressed because they are unfavourable. Personal and institutional information should be protected according to applicable requirements.

Finance officials should participate before policy adoption so that cost and delivery assumptions can be tested. Budget documents should distinguish development, transition and recurrent expenditure. Transfers to subnational bodies or institutions should have a defined purpose while allowing reasonable local management.

Audit should examine whether funds were authorised, received, used for the stated purpose and supported by records. It should not infer educational impact from financial compliance. Educational evaluation and financial audit answer different questions and should exchange evidence without confusing their mandates.

Procurement should preserve quality, competition and continuity. Specifications for materials or equipment should derive from curriculum need. Contracts should address delivery, translation or accessibility where relevant, maintenance, ownership and protection from commercial use of learner information.

An advisory group may include educators, learners, communities, researchers, public-sector specialists and other contributors. Membership should reflect the questions considered and should disclose relevant interests. The group should advise; domestic authority remains with the designated public bodies.

Partnership agreements should identify purpose, contribution, intellectual control, cost, duration, evidence, safeguarding and exit arrangements. A partner's reputation does not replace review of materials or claims. Institutions should not imply endorsement beyond the agreement.

Learner participation should be supported through accessible information and safe processes. Learners should know how their views will be used and should receive feedback. Participation should not require disclosure of personal circumstances unrelated to the decision.

The operating model should establish where institutions report contradictions or unmet needs. A curriculum question should reach the curriculum authority; a funding shortfall should reach the responsible finance body; a safety concern requires immediate handling under applicable procedures. A single general reporting channel may receive matters but must route them to an owner.

Resolution records should identify the issue, evidence, affected population, interim protection, responsible body, decision and review date. Repeated local failures should trigger examination of national design. The system should not close a case merely because guidance was reissued.

Where authorities disagree, an escalation route should identify the final decision-maker. Public reporting should acknowledge material unresolved issues without exposing confidential information. Transparency about a continuing constraint is preferable to an unsupported claim of completion.

Source: ICEQC operating model for this report.

The operating cycle should connect evidence with decisions. At the start of a planning period, authorities review previous findings, unresolved actions and policy changes. Curriculum and professional bodies identify required support. Finance confirms available resources and allocation. Statistical and research bodies confirm the evidence programme.

During implementation, routine monitoring should focus on material delivery conditions and risks. Institutions and local authorities should receive feedback, not only requests for information. Mid-year review may correct distribution, guidance or timetable while change is still possible.

At the end of the period, bodies should report against assigned responsibilities. The national report consolidates findings, limitations, expenditure and corrective actions. The next cycle begins from those decisions. An annual cycle does not require every outcome measure every year; frequency should match the rate of change and burden.

Implementation is institutionalised when responsibilities, resources, professional capability, evidence and corrective action operate through ordinary systems without dependence on exceptional individuals or temporary projects. The presence of a permanent committee alone does not establish institutionalisation.

Evidence should show that curriculum and assessment documents are current, new teachers receive preparation, serving teachers can access support, institutions include the entitlement in ordinary planning, data are produced on schedule, finance reaches responsible levels and findings lead to action.

Institutionalisation does not imply that policy is complete. The subject and evidence will develop. A mature system can revise its arrangements while preserving authority, records and public accountability.

The first research priority is to describe curriculum opportunity across systems and learner groups. Existing policy documents can show formal entitlement, but studies of taught content, learner work and assessment are needed to determine what learners encounter. Research should examine variation by phase, pathway, location, language and institutional resources.

Studies should avoid beginning with a fixed list of sustainability words. They should code the educational substance: knowledge, reasoning, application, participation and progression. The relationship between local examples and wider systems requires particular attention.

Research should identify the knowledge and support teachers require in different phases and subjects. Surveys can describe confidence and participation, but observation, planning evidence and learner work are needed to examine practice. Studies should distinguish lack of content knowledge, pedagogical uncertainty, contradictory assessment, insufficient time and material constraints.

Professional-development research should record programme intensity, applied components, trainer preparation, cost and follow-up. Evaluations should compare delivery models under ordinary conditions and should examine reach to remote and disadvantaged teachers.

There is no settled international measure of education for sustainable development in 2005. Research should define bounded constructs and test tasks before pursuing comparison. Possible domains include interpretation of evidence, systems reasoning, consideration of distribution and justified decision-making.

Studies should examine language demand, cultural familiarity, curriculum opportunity and the effect of different response formats. Attitude measures require caution because responses can reflect social desirability and do not establish knowledge or action. Behaviour measures should not penalise learners for household constraints.

Research should investigate how leadership, professional collaboration, learning time, materials, facilities and participation affect implementation. These conditions interact; studies based on one input may omit the mechanism through which it matters.

Institutional operations should be examined in relation to learning. Research can test whether use of the institution as a practical setting improves understanding, which learners participate and what support is required. Operational improvement and educational outcome should remain separate variables.

Cost studies should identify the full resources required by different models, including teacher time, translation, remote delivery, accessibility, maintenance and data. Pilot expenditure should be separated from recurrent cost. Research should compare allocation and actual use.

Equity analysis should examine whether disadvantaged institutions receive additional support and whether the support is adequate to produce the common entitlement. Cost-effectiveness should be assessed within minimum quality and equity constraints.

Research should examine whether curriculum, assessment, teacher policy, finance and accountability support one another. Comparative policy studies can identify different arrangements without ranking them prematurely. Case studies should explain authority, sequence and conditions.

Implementation failures provide evidence. Studies should examine how systems detect contradictions, route concerns and revise policy. A system capable of correction may be stronger than one reporting uninterrupted compliance from weak evidence.

The Decade invites interest in effects beyond educational attainment, including participation, occupational practice and public decision-making. Such research requires longitudinal design and careful treatment of selection and context. Education is one influence among many.

Researchers should resist using correlation with schooling or programme participation as proof of effect. Intermediate outcomes—knowledge, reasoning, capability and participation within defined settings—provide a more defensible chain. Wider effects can be studied where measures, timing and comparison support the inference.

International cooperation can improve definitions, instruments and comparative understanding. It should begin with shared metadata and transparent methods. Countries should retain control of national data according to applicable arrangements and should receive usable findings.

Collaborative studies should include systems with different languages, income levels and administrative structures during design, not only as data providers. Instruments developed in one context require adaptation and renewed validity evidence elsewhere.

Publication should identify participating populations and limitations. A study covering a group of countries should not be described as global unless the design supports that scope.

Research involving learners should have lawful authority or appropriate consent arrangements, protect confidentiality and minimise burden. Participation should not affect educational access or assessment unless the study has that authorised purpose and adequate safeguards.

Findings should be published even when they do not support the intervention. Funding and interests should be disclosed. Communities that contribute knowledge or data should receive an account of findings in an accessible form.

Historical evidence should remain dated. Later research may show that an early assumption was wrong; the correction belongs in subsequent work, not in silent revision of the 2005 record.

Source: ICEQC research programme for this report.

This appendix provides an analytical framework for authorities reviewing the treatment of education for sustainable development within national quality policy. It is not a model law, curriculum or inspection instrument. Its purpose is to make the reasoning of the main report reproducible and to support comparison among policy components without collapsing legitimate national differences.

The review begins with documentary evidence but does not end there. A policy text establishes an authorised expectation; implementation evidence establishes whether institutions and learners encounter it. Reviewers should retain the distinction and record the strength of evidence for each finding.

Source: ICEQC analytical framework derived from the sources and tests set out in this report.[REF-01] [REF-05] [REF-06] [REF-08]

The framework does not require a numerical score. Reviewers should record one of four evidentiary positions for each domain:

• **Not established**: no relevant authorised expectation or evidence is available.
• **Formally established**: an authorised expectation exists, but implementation evidence is absent or insufficient.
• **Partly evidenced**: implementation evidence exists for some settings, populations or elements, with material gaps or uncertainty.
• **Substantially evidenced**: multiple sources show that the expectation is implemented across the defined scope, with limitations and distributional differences disclosed.

“Substantially evidenced” should not be interpreted as complete or uniform. It indicates that the available evidence is sufficient for the stated conclusion and scale. A finding should identify the period and population covered. Where evidence is contradictory, the rating should reflect the weaker position unless the difference can be explained by scope or method.

The mandate, coherence, feasibility and equity tests should be applied to each proposed measure.

Under the **mandate test**, reviewers identify the domestic authority and the international or national purpose relied upon. They examine whether the measure remains educational in character and whether it respects rights, professional responsibilities and legitimate differences in viewpoint.

Under the **coherence test**, reviewers trace the measure through curriculum, assessment, teacher policy, institutional review, finance and reporting. A contradiction should be recorded even where each component is individually well drafted. For example, a curriculum expectation requiring field investigation is not coherent if assessment, timetable and safety arrangements make such investigation unavailable to most learners.

Under the **feasibility test**, reviewers examine resources, professional capability, implementation time and administrative burden. Feasibility is not a reason to remove the entitlement of learners in disadvantaged settings. It is a reason to adapt the means, allocate additional support or extend the timetable while maintaining public responsibility.

Under the **equity test**, reviewers identify groups, locations and pathways likely to receive different opportunities or bear different costs. They consider the quality of disaggregated data and the risk that averages or voluntary participation conceal exclusion.

Every indicator used for national reporting should be accompanied by metadata sufficient for another competent analyst to reproduce or interpret it. The metadata should state:

• the policy question the indicator addresses;
• the exact definition and unit;
• the population and exclusions;
• the numerator and denominator for a rate or proportion;
• the data source and collection method;
• the reference date or period;
• the frequency of collection;
• the level of geographic and population disaggregation;
• rules for missing, suppressed or estimated values;
• known threats to validity or comparability;
• the body responsible for production and quality control; and
• the date and nature of any definition change.

Indicators should be reported as a profile. No weighting system is proposed. The evidence available at the commencement of the Decade does not support a universal composite index, and the policy consequences of different weights would require public justification.

Source: ICEQC indicator design for this report. The indicators are proposals for national adaptation, not internationally agreed measures.

A denominator should represent the population to which the policy claim refers. If the policy applies only to public lower-secondary schools, reporting should not imply coverage of all learners. Where enrolment data are used to weight institutional results, the reference year and treatment of learners in multiple programmes should be stated.

Missing data should be classified. At minimum, reporting should distinguish not applicable, not collected, not reported, suppressed for protection, and invalid. These categories have different meanings and should not be combined with zero. The proportion of the intended population represented by valid data should accompany every principal indicator.

Imputation may be appropriate for statistical analysis but should not be undertaken solely to complete a public table. The method, assumptions and sensitivity of findings should be disclosed. For early monitoring, transparent incomplete data may be more credible than a complete estimate built on weak assumptions.

Disaggregation should respond to a policy question. Categories need stable definitions and sufficient sample or population size. Publication should not permit identification of individual learners or expose small groups to harm. In small systems, multi-year aggregation or broader geographic grouping may be necessary, but the effect on timeliness and interpretation should be stated.

Differences should be presented with their magnitude and uncertainty where sampling is used. Statistical significance alone does not establish educational importance. Conversely, a difference affecting a small or marginalised population should not be dismissed solely because a national aggregate is unchanged.

The documentary review should include the current national statement of educational aims, curriculum frameworks, subject specifications, assessment requirements, teacher standards, initial teacher-education requirements, continuing professional-learning guidance, institutional review criteria, education-sector plan, relevant budget documents and official reporting requirements. Superseded documents may be required to establish the baseline and timing of change.

Each document should be recorded with title, issuing body, authority, status, adoption date, effective date, scope and language. Draft and adopted documents should be stored separately. Where different regions use different frameworks, the sampling and legal relationship should be described.

Curriculum coding should examine substance rather than keywords. A coding manual should define:

• whether the expectation concerns knowledge, skill, judgement, participation or institutional activity;
• the relevant social, economic and environmental dimension or relationship;
• the level of cognitive demand;
• whether the expectation is required, recommended or illustrative;
• the educational phase and pathway;
• the accountable subject or programme;
• whether assessment evidence is specified; and
• whether progression from an earlier or to a later stage is visible.

At least a sample of documents should be coded independently by two reviewers. Differences should be discussed and the coding manual revised. Agreement statistics may be reported where the volume and purpose justify them, but numerical agreement should not replace examination of substantive disagreement.

An institutional study should include variation relevant to implementation: urban and rural location, resource level, language, institution size, public administrative region and education pathway. A purposive component may be used to understand unusual conditions, but national estimates require probability sampling or a clearly justified equivalent design.

Evidence may include timetables, schemes of work, learner work, assessment tasks, professional-learning records, resource inventories, minutes of relevant decisions, interviews and observation. No single source should determine the finding where corroboration is feasible. Staff and learner accounts are evidence of experience and interpretation; they should not be dismissed because they differ from documents.

Learner tasks should be linked to taught content. A task may provide a short body of evidence and ask learners to identify a question, interpret relationships, distinguish evidence from assertion, compare consequences and justify a feasible response. Scoring criteria should be published in sufficient detail for moderation.

Tasks should avoid requiring personal disclosure about household behaviour or beliefs. Contextual information may be collected for analysis under appropriate protections, but assessment performance should not depend on a learner's access to resources outside the institution.

Findings should distinguish:

• **documented expectation** — what policy or curriculum requires;
• **reported practice** — what participants say occurs;
• **observed opportunity** — what review evidence shows learners encountered;
• **demonstrated capability** — what learners show on defined tasks; and
• **inferred explanation** — the review team's interpretation of factors contributing to the finding.

This distinction should appear in tables and narrative conclusions. It reduces the risk that an inference is repeated later as recorded fact.

Source: ICEQC risk analysis for the first phase of national implementation.

The register should be adapted to domestic authority and reviewed at defined intervals. A listed control is not effective because it exists on paper. Authorities should examine whether the control operated, whether it detected the risk and whether corrective action followed.

The matrix below illustrates how a national curriculum authority may describe progression without prescribing identical content. It is organised by capability rather than school subject. Each jurisdiction should map the capability to its curriculum structure, languages and educational stages.

Source: ICEQC analytical progression designed for this report. It is informed by the broad educational purposes in the source instruments but is not an adopted international curriculum.[REF-01] [REF-09] [REF-10]

Curriculum authorities should first identify existing expectations that correspond to the capabilities. The mapping should record exact locations and should not infer coverage from general preambles. Where several subjects contribute, the authority should determine whether their combined progression is coherent and whether learners in all pathways receive the opportunity.

The stages are descriptive rather than age-fixed. Systems should adapt them to the structure and prior learning of their population. Movement to a later description presumes the knowledge and language necessary for the task. A sophisticated discussion unsupported by subject knowledge does not demonstrate the capability described.

The matrix should not be converted directly into a checklist for every lesson. It informs curriculum design, sample assessment and programme review. Teachers require freedom to sequence learning in ways appropriate to subject and learners, within the national entitlement.

Reviewers should ask whether:

• expectations increase in intellectual demand rather than repeat general messages;
• later stages require stronger evidence and treatment of uncertainty;
• local examples develop towards national and global relationships;
• learners encounter both knowledge and opportunities to apply it;
• progression is available across academic, vocational and other pathways;
• language demand is appropriate and supported; and
• assessment samples the intended capability without reducing the curriculum to the test.

Before developing tasks, the authority should state the capability to be assessed and the inferences intended. “Sustainability awareness” is too broad for a defensible assessment. A more precise construct might be: the capacity to interpret a defined body of evidence about a local resource question, identify consequences for affected groups, and justify a feasible response while acknowledging uncertainty.

The construct statement should identify knowledge assumed, reasoning required and content outside scope. It should explain whether the result is intended for classroom feedback, programme evaluation or reporting at system level. An instrument suitable for one use may not be suitable for another.

A task should provide sufficient information for learners who received the relevant curriculum. It may include a short text, map, diagram, table or description of alternatives. Extraneous reading demand should be limited unless literacy itself forms part of the construct. Visual material requires accessible labels and alternatives for learners who cannot use it in the same way.

Prompts should permit more than one defensible conclusion where the issue genuinely involves judgement. Scoring should distinguish use of evidence, recognition of consequences, feasibility and quality of justification. Agreement with a designated option should not receive credit independently of reasoning.

Tasks involving local knowledge should not assume that every learner has identical experience. Information essential to the response should be supplied or clearly established through the taught programme. Personal or household behaviour should not be required for evidence.

Source: ICEQC assessment framework for this report.

Tasks should be reviewed by subject specialists, teachers, assessment specialists and experts in language and inclusion. Cognitive interviews or structured learner discussion may identify misunderstanding of instructions. A pilot should include learners from varied settings and record non-response, time, accessibility and patterns of unexpected interpretation.

Scoring guidance should include examples at different levels and explanations of borderline decisions. Two or more scorers should mark a sample. Disagreement should inform revision of criteria and training. Reliability evidence should be reported, but high agreement does not establish that the task measures the intended construct.

Validity review should examine curriculum opportunity, response processes, scoring consistency, relationships with other relevant evidence and possible consequences of use. Differential performance among groups requires investigation of both real opportunity differences and task bias. No high-stakes consequence should attach to an experimental measure.

System-level results should report the sampled population, participation, exclusions, weighting, uncertainty and reference period. Performance levels should be described in terms of demonstrated capability, not labels of institutional virtue. Trend reporting should wait until tasks or scales are demonstrably comparable.

The assessment should remain one source within the wider quality profile. It cannot establish institutional practice, teacher capacity, finance or the full educational aims of the Decade.

Implementation costing should identify development, transition and recurrent costs. Development costs may include policy analysis, curriculum design, consultation, materials and assessment research. Transition costs may include professional learning, replacement of materials, local planning and pilot evaluation. Recurrent costs may include continued teacher support, data collection, moderation, maintenance and additional instructional resources.

Professional time should be valued even where it does not create a new cash payment. Unrecorded workload is a real implementation cost and may affect retention or other educational duties. Costing should identify which body and level bear each cost.

Unit costs should use a denominator suited to the activity: cost per teacher prepared, institution supported, learner covered or assessment completed. The calculation should state whether central development costs are included and over what period they are spread. Pilot unit costs may fall or rise at scale; assumptions should be explicit.

Remote delivery, translation, accessibility and small-institution costs should be modelled separately where material. A single national average may understate the resources required for equitable coverage. Authorities should distinguish cost differences arising from legitimate delivery conditions from inefficient procurement or administration.

Expenditure should be compared with the minimum resources required by the implementation design. A fair distribution of an inadequate total does not establish feasibility. Conversely, greater expenditure does not establish quality if funds do not purchase relevant support or reach institutions.

Distributional analysis should compare allocation and actual expenditure across nationally relevant groups. It should examine per-learner amounts, fixed costs, need and service availability. Ratios should be interpreted with context: a remote institution may appropriately require higher expenditure per learner.

The plan should identify dependence on time-limited grants, external assistance, voluntary labour or price-sensitive inputs. Recurrent obligations should not be created without a credible funding source. Technology and capital investments require maintenance, replacement and training costs.

Where fiscal conditions deteriorate, authorities should establish priorities in advance. Protection of safety, foundational provision and minimum curriculum entitlement should take precedence over promotional activity and non-essential capital purchases.

Cost-effectiveness analysis compares alternative means of achieving a defined educational result. It should not assign monetary value to every outcome. For example, authorities may compare delivery models by cost per teacher who completes preparation and demonstrates specified capability. The outcome definition, period and uncertainty should be stated.

Comparisons should avoid selecting a cheaper model that reaches only accessible institutions or produces a narrower outcome. Equity and quality constraints belong within the decision rule.

A national progress report should contain:

1. the domestic mandate, scope and status of relevant instruments;
2. the reporting period and evidence cut-off;
3. the implementation model and responsibilities;
4. the baseline and material changes since it was established;
5. capacity, opportunity, outcome and equity findings presented separately;
6. expenditure and distributional information;
7. results of evaluations, including negative and unintended effects;
8. data definitions, coverage, missingness and limitations;
9. corrective actions, responsible bodies and review dates; and
10. a reference register identifying the evidence relied upon.

Material public claims should be linked to the evidence supporting them. A claims register may be maintained as a technical control, even if not published in full. It should record the claim, scale, period, source, responsible analyst, limitation and approval. The visible report should state the material limitation near the claim rather than requiring readers to infer it from a distant annex.

Every table should identify title, unit, reference period, population and source. Percentages should state the denominator in the title or note. Totals and components should reconcile or the reason should be explained. Rounding should not create apparent precision beyond the data.

Figures should not truncate axes in a way that exaggerates difference. Maps should state classification rules and treatment of missing areas. Colour should not be the sole means of conveying meaning. Narrative should identify the finding without repeating every value.

Profiles should follow a common evidence core while allowing context-specific analysis. A profile should not be reduced to a scorecard. It should explain authority, resources, implementation stage, principal findings and corrective priorities. Comparison among profiles should respect differences in scope and data.

Institutions should not be named in public reports on sampled quality evidence unless authority, method and fairness safeguards permit it. The primary purpose of early studies is system learning. Serious safety or rights concerns require action through the appropriate process and should not be concealed by research confidentiality.

The main findings should be available in forms and languages accessible to affected populations. Technical precision should be preserved through a full report and annexes, while summaries explain the result without promotional simplification. Translation should be reviewed by subject and language specialists; key definitions should remain consistent.

Corrections and revisions should be dated and described. A statistical correction should identify affected tables and conclusions. Changes to policy status or evidence after the cut-off belong in a later edition or report, not in a silent alteration that makes the historical publication appear to contain later knowledge.

The baseline should answer five questions. What authorised policy exists? What curriculum opportunity is specified? What professional and institutional capacity is available? What evidence shows that learners receive the opportunity? Which populations and locations are absent or disadvantaged in the evidence?

Each question should be connected to a source and method. Legal and policy status requires documentary verification. Curriculum opportunity requires structured review. Capacity requires administrative and sampled institutional data. Learner opportunity requires evidence from practice. Equity requires denominators and disaggregation.

The documentary component covers national and, where relevant, subnational law, policy, curriculum, assessment, teacher requirements, finance and institutional-review documents. The administrative component uses enrolment, staffing, professional development, material distribution and expenditure data with existing definitions reviewed. The institutional component samples settings for direct evidence. The learner component examines work and experience without introducing premature high-stakes testing.

Components should be connected by stable institutional identifiers where lawful and technically sound. The design should not assume that records collected for administration are accurate enough for research. Definitions, missingness and incentives affecting reporting require examination.

The institutional sample should be designed from the claims intended. A national estimate requires a probability basis and appropriate weighting. A diagnostic study may use stratified purposive selection but should not report national prevalence. Relevant strata can include region, urban or rural location, institution type, size, language and a nationally recognised measure of disadvantage.

Replacement of non-participating institutions can bias findings. The protocol should define contact, follow-up and replacement rules and should report response rates by stratum. Reasons for non-participation may themselves indicate capacity constraints.

Learner sampling should reflect the phase and pathways covered by policy. Exclusions should be minimised and documented. Accessibility arrangements should preserve the construct under review.

The study may use a document coding form, head-teacher questionnaire, teacher questionnaire, learner questionnaire, resource inventory, observation schedule, work-sample protocol and interview guides. Each instrument should have a defined role; collecting the same self-report through several forms adds burden without necessarily strengthening evidence.

Questions should avoid socially desirable phrasing. Asking whether a school “supports sustainability” invites affirmation. Asking which curriculum sequence was taught, which materials were used and what evidence is retained produces more interpretable information.

Observation should record defined features and context. It should not assign a broad quality judgement from a single lesson. Work-sample coding requires criteria, reviewer preparation and checks on agreement.

Fieldworkers should receive training on purpose, definitions, consent or other applicable authority, neutrality, safeguarding and data handling. Translation should be piloted. Supervisors should review completeness and a sample of evidence while fieldwork is under way.

Institutions should know the scope and use of the study. Participation in research should not be presented as a certification or guarantee of favourable judgement. Serious safety or rights concerns should follow an approved escalation process.

Electronic collection may reduce entry error where infrastructure is reliable. It should not exclude remote sites or create a different instrument by location without equivalence review. Paper records require controlled entry and verification.

Analysis should preserve weights, clustering and stratification where the design requires them. Estimates should include uncertainty. Qualitative evidence should be coded through a documented process and used to explain mechanisms and variation.

The baseline report should publish instruments or sufficient descriptions, response, missingness, limitations and the evidence cut-off. Datasets should be archived with access and protection appropriate to their content. A public-use file should not permit identification.

Findings should be reviewed for consistency with their source. A statistical association should not be rewritten as causal explanation. An institutional example should not be introduced as a national pattern. Editorial review should verify that qualifications survive into the executive summary.

Every source should have a stable identifier, title, authoring body, type, date, status, URL or archive location, relevant pages or provisions and evidence role. Drafts and final instruments should have separate records. A later digital copy of an earlier instrument should preserve the instrument's original date while recording the provenance of the copy internally.

The register should distinguish adoption, publication, effective and access dates. These dates answer different questions. A resolution may be adopted before the official document is distributed; a law may be published before it takes effect; a dataset may be released after its reference year.

A citation supports only the proposition warranted by the source. A legal instrument can establish that a commitment was adopted, but it does not establish implementation. A monitoring report can provide empirical analysis, but it does not create domestic authority. A comparative assessment supports conclusions within its population and construct.

Material numerical claims should identify the table, page or series where feasible. Calculations should identify source values and method. If a figure is estimated, the report should not describe it as observed.

A source may support context without supporting the recommendation drawn by the report. The text should signal the analytical step. Citation density alone does not establish evidentiary quality.

The evidence cut-off should be applied to publication and institutional status. Later evidence should not be used to improve the apparent prescience of a historical report. Where an earlier instrument is accessed through a later archive, analysts should verify the original text and date and should not rely on later commentary for a contemporaneous conclusion.

Sources without a reliable date should not support time-sensitive claims. If retained for a non-temporal definition, that limited role should be recorded. A later edition that reproduces an earlier statement should not replace the earlier source where the historical availability of the statement matters.

If a source error or citation mismatch is identified after publication, the correction record should identify the affected statement, original source, corrected source or wording, date and effect on conclusions. The authoritative files and hash should be revised through a new controlled version. The original historical publication should remain archived.

Minor typographical correction should not be used to introduce later evidence. Substantive revision requires a new edition or clearly labelled corrigendum.

Citation of an international body does not imply its endorsement of this report. The report should distinguish source findings from ICEQC analysis. Language that attributes a national policy recommendation to a source should be used only where the source makes that recommendation within the relevant scope.

Reference lists should not be enlarged for appearance. Every source should perform an identifiable role, and principal claims should rely on the most authoritative and direct source available by the cut-off.

The following definitions control the meaning of terms used in this report. They are analytical definitions for the present study unless a source is identified. They do not create international or domestic legal definitions.

**Baseline** — the documented condition at a stated reference point against which later comparable evidence may be examined. A baseline records prior activity where it exists and does not imply that implementation began at zero.

**Capability** — an evidenced capacity to use relevant knowledge, skill and judgement for a defined educational task. Capability is narrower than a claim about habitual conduct in all settings.

**Capacity indicator** — a measure of the personnel, knowledge, time, material, finance or organisational conditions available for implementation. It does not by itself establish that learners received the intended opportunity.

**Common entitlement** — the minimum educational opportunity that all learners within a defined population should receive, while allowing contextual variation in examples and methods.

**Comparability** — the extent to which results have sufficiently consistent definitions, populations, reference periods and methods for the comparison proposed. Similar labels do not establish comparability.

**Completion** — fulfilment of a defined programme or education stage according to published rules. Completion should not be inferred from enrolment or attendance alone.

**Contextual adaptation** — modification of content, example, language, delivery or support to reflect local conditions while preserving the intended educational purpose and demand.

**Corrective action** — an assigned and time-bound response to an evidenced failure or risk, including the change required, responsible body, resources, evidence of completion and follow-up.

**Curriculum entitlement coverage** — the proportion of learners or curriculum pathways for which a specified learning opportunity is formally required. It is different from evidence that the curriculum was taught.

**Disaggregation** — separation of data into policy-relevant groups or locations to examine distribution. Categories require clear definitions, adequate quality and protection against harm or identification.

**Education for sustainable development** — education that enables people to understand and address connected social, economic and environmental conditions affecting present and future well-being. In this report, the term concerns educational purpose and opportunity rather than adherence to a single policy programme.

**Educational opportunity** — the content, teaching, time, resources and participation through which a learner can develop the stated capability. A written curriculum is one condition of opportunity, not proof of its delivery.

**Effective date** — the date from which an instrument or requirement applies. It may differ from adoption or publication.

**Enabling input** — a resource or condition supporting learning, such as capable teachers, instructional time, materials or safe facilities. Its presence should not be reported as an outcome.

**Equity** — fair distribution of educational opportunity, resources and outcomes with attention to differing needs and barriers. Equal nominal allocation may be inequitable where delivery costs or starting conditions differ.

**Evidence cut-off** — the latest date on which evidence or institutional status may enter the analysis. Later knowledge is excluded from the report's contemporaneous conclusions.

**Foundational capability** — literacy, numeracy and other essential knowledge or skill enabling further learning, interpretation and participation. The term does not imply that relevant applied learning must wait until a separate foundational phase is complete.

**Implementation** — operation of an authorised policy or programme within a defined scope, evidenced beyond adoption. The necessary evidence depends on whether the claim concerns responsibility, capacity, institutional practice or learner opportunity.

**Indicator specification** — the controlled definition of a measure, including purpose, population, unit, source, period, disaggregation, missing-data rules, limitations and responsibility.

**Institutionalisation** — incorporation of responsibility, resources, capability, evidence and corrective action into ordinary systems so that implementation does not depend on exceptional individuals or temporary projects.

**Learning outcome** — knowledge, skill, capability or other educational result demonstrated through evidence appropriate to the construct. A learning outcome is not identical to a wider social effect.

**Monitoring** — repeated, defined collection and interpretation of evidence to understand status, distribution and change. Monitoring can identify patterns but does not by itself establish causation.

**National quality policy** — the public framework through which authorities define educational purposes and expectations, organise and support provision, review performance and account for results.

**Opportunity-to-learn evidence** — evidence that learners encountered the content and conditions relevant to an assessment or policy claim. It supports interpretation of outcome differences.

**Outcome indicator** — a measure of a defined result. Its interpretation depends on construct validity, population, opportunity and the use for which it was designed.

**Participation** — a process through which learners, communities or other affected groups contribute information or views to learning or decision-making. Participation requires transparent scope and does not automatically transfer decision authority.

**Pilot** — a bounded implementation used to test feasibility, validity, cost or effect before a decision on wider use. A pilot should have a protocol, defined population, evaluation and decision point.

**Policy coherence** — consistency and functional alignment among educational purpose, curriculum, assessment, teacher policy, finance, institutional responsibilities and reporting.

**Progression** — an increase in the knowledge, reasoning, scale, independence or integration expected of learners across stages. Repetition of topics without greater demand is not progression.

**Proportionality** — alignment of requirements, evidence and intervention with purpose, risk, scale, capacity and consequence. Proportionality does not permit reduction of rights or minimum entitlement.

**Public-awareness provision** — learning or communication directed to public understanding outside or alongside formal education. Audience reach is activity evidence, not proof of learning.

**Quality profile** — presentation of distinct dimensions of conditions, opportunity, outcomes and equity without concealing weakness through a single composite score.

**Research evidence** — evidence produced through an explicit question, method and analysis. Research can inform public decisions but does not acquire regulatory authority solely through methodological independence.

**Source status** — the institutional and procedural position of a source, such as treaty, adopted resolution, final policy, draft document, monitoring report or study. Status determines the claims the source can support.

**Sustainable development** — a policy concept involving relationships among social, economic and environmental conditions and consequences over time. This report does not prescribe a single national model or resolve all contested policy choices carried by the term.

**Taught opportunity** — evidence that the intended curriculum was delivered to learners, examined through records, work, observation or reports of sufficient quality. It is more specific than formal curriculum coverage.

**Validity** — the strength of evidence and reasoning supporting the interpretation and use of a measure. An instrument is not valid in the abstract; validity concerns proposed conclusions and consequences.

Public reports should use the most specific term warranted. “Policy adopted” should not become “implementation completed”. “Participants reached” should not become “learners trained” without evidence. “Improved score” should not become “improved sustainability” where the measure covers a bounded educational capability.

Definitions should be stable across a reporting period. If a term changes, the report should identify the change and its effect on earlier figures or conclusions. Translation should preserve controlled meaning rather than literal form alone.

This report is a policy synthesis at the commencement of the Decade. It does not measure national implementation, estimate global cost or establish a comparative performance scale. The available evidence does not permit those conclusions.

The international record provides broad direction across different forms of authority. Its application depends on domestic law, administration, curriculum and finance. The report therefore identifies national policy controls rather than a universal legal procedure.

Comparative outcome evidence is limited in country, population and construct coverage. The discussion of PISA 2003 is methodological and does not represent participating countries as having implemented the Decade.[REF-11]

The 2004 International Implementation Scheme remains a draft at the evidence cut-off. Its concepts are used for preparatory analysis only. Later consideration and any final text fall outside this report.[REF-05] [REF-07]

Regional sections identify planning conditions and risks; they are not regional performance assessments. Diversity within regions prevents uniform conclusions. Country-level decisions require country evidence.

The proposed indicators, matrices, cases and protocols have not been internationally validated. They are transparent analytical tools for adaptation and testing. Their inclusion should not be interpreted as adoption by any cited organisation.

The evidence cut-off excludes later data and experience, even where they could clarify the subject. This limitation is deliberate: the report represents what can responsibly be stated on 9 February 2005.

A national decision record should preserve the basis on which an authority acts. It supports continuity, later evaluation and public accountability. The record is not a substitute for the authorised instrument; it explains the evidence, alternatives and implementation controls underlying it.

The record should state the decision title, responsible body, legal or administrative authority, decision date, effective date, scope, affected education phases and relationship to existing plans. Draft, approved and superseded status should be unambiguous.

The international sources considered should be listed with their status. The record should explain which elements informed the decision and should avoid suggesting that a national choice was approved by a cited organisation.

The problem statement should identify the observed condition, population, period and evidence. It should separate empirical findings from policy judgement. Gaps and uncertainty should remain visible.

Evidence should include the baseline, curriculum mapping, implementation capacity, resource position, equity implications and consultation findings where relevant. Each material claim should be traceable to the source register.

The authority should record feasible options, including continuation of current arrangements where it is a genuine alternative. Appraisal should examine educational value, rights, coherence, feasibility, cost, distribution, administrative burden and risk.

An option should not be rejected solely because it requires resources if the existing arrangement fails to provide the minimum entitlement. The decision should identify how resources will be secured or how implementation will be phased without concealing the gap.

The record should describe the selected measure in operational terms: expected result, responsible bodies, required actions, timetable, finance, support, evidence and review. It should identify which elements are mandatory, which are guidance and which remain under pilot.

Dependencies should be stated. A curriculum change may depend on teacher preparation and assessment revision; an outcome study may depend on sufficient exposure and a valid instrument. The schedule should reflect those relationships.

The decision should identify groups or locations at risk of reduced access, higher burden or weaker evidence. Mitigation should have an owner and resource. Safeguarding, accessibility, language and protection of personal information should be addressed according to the measure.

External contributions should be disclosed. Agreements should preserve public responsibility, curriculum control and continuity of the learner entitlement.

The record should list indicators and other evidence with definitions, baseline, frequency and responsible producer. It should state the claim each source can support. A review date and decision rule should be established before results are available.

Corrective-action routes should identify how institutions and local authorities report infeasible or contradictory requirements. The central authority's response should form part of subsequent public reporting.

The public version should contain sufficient information to understand the decision and its evidence. Confidential or personal material may be withheld under applicable arrangements, but this should not remove the basis for material public claims.

The record, source register and approved instrument should be preserved together. Later amendments should identify the earlier decision and explain what changed. Historical versions should remain available so that progress can be assessed against the policy actually in force at the time.

A decision record is complete when another competent official can determine: who decided; under what authority; what problem and population were addressed; what evidence and alternatives were considered; what duties, resources and timetable were approved; how equity and risk were handled; what evidence will be produced; and when the decision will be reviewed.

Completion of the record does not establish successful implementation. It establishes that the decision is explicit, traceable and capable of later scrutiny.

Before approval, the responsible body should answer the following questions in the record:

1. What educational problem is established by the evidence, and which learners, institutions or education phases are within scope?
2. Which domestic authority permits the proposed action, and which body remains accountable for its implementation?
3. Which international instruments or reports inform the decision, what is their status, and which propositions do they support?
4. How does the measure relate to existing Education for All, curriculum, teacher, assessment and national-development commitments?
5. What knowledge, capability or educational opportunity will change, and how will that change become visible in ordinary provision?
6. What alternatives were considered, including revision of existing arrangements rather than creation of a parallel programme?
7. What teacher knowledge, professional time, materials, leadership and administrative capacity are required?
8. What are the development, transition and recurrent costs, which level bears them, and how will disadvantaged or high-cost settings be supported?
9. Which examinations, accountability rules or institutional incentives could obstruct the intended change, and how will they be aligned?
10. Which groups may receive a weaker opportunity or a higher burden, and what evidence will show whether mitigation operates?
11. Which elements are adopted requirements, which are guidance, which are pilots and which remain proposals?
12. What baseline exists, what new data are necessary, and which claims can each measure support?
13. How will missing data, small populations, language differences and changes in definition be handled?
14. What adverse or unintended effects will be monitored, and what interim protection applies where risk is material?
15. How can institutions report a contradiction, resource failure or safeguarding concern, and which body must respond?
16. What evidence would justify continuation, revision, expansion or discontinuation, and when will that decision occur?
17. What information will be published, in which languages and forms, and how will technical limitations remain visible?
18. Which records and versions will be preserved so that later reviewers can reconstruct the decision as it stood in 2005?

An unanswered question should not always prevent approval. The record should state why the uncertainty is acceptable, how it will be investigated and which action depends on its resolution. This treatment converts uncertainty into a controlled research or implementation question rather than allowing it to disappear from the policy account.

Approval should also identify the evidence that would invalidate a material assumption. If the selected delivery model depends on teacher access to a resource, the review should define what level of non-receipt requires redesign. If an assessment assumes a taught curriculum, the result should not be interpreted until opportunity-to-learn evidence reaches the stated coverage. If a funding formula assumes that local authorities can meet a residual cost, the authority should test that assumption across low-revenue areas before describing the entitlement as nationally financed.

These conditions prevent a decision from becoming self-confirming. Evidence is not collected only to record activity under the approved measure; it is used to test whether the measure remains educationally justified, feasible and equitable. The approval record should therefore authorise revision as well as implementation and should state who may initiate that revision when evidence shows material failure.