The research, published in Discover Education by Ashiya Osman, Athenkosi Sigenu and Charmaine du Plessis of the University of South Africa, analysed 31 completed projects from the Department of Basic Education’s Youth Citizen Action Programme (YCAP), a national initiative in which teams of four to five learners identify a community problem, investigate its causes, implement an intervention and reflect on the results. The portfolios, produced between 2021 and 2024 by primary and secondary school teams across several provinces, were treated by the researchers as socially produced texts — artefacts of learning shaped by interactions with peers, mentors, communities and the programme’s structured toolkit.

The theoretical backbone of the study is a pairing of two frameworks. The first is Wiek and colleagues’ influential model of sustainability competencies, which singles out systems thinking, normative thinking and strategic thinking as core capacities for anyone hoping to act on complex social and environmental problems. The second is variation theory, developed by Ference Marton and colleagues, which holds that learning happens not merely through experience but through discernment — through noticing how certain aspects of a situation vary against a stable background. Applied to civic education, the theory suggests that the sheer diversity of community challenges learners encounter is not a complicating variable but the very mechanism that makes deep learning possible.

The researchers adopted a qualitative design grounded in social constructivism, purposively sampling portfolios that had been recognised at provincial or national level and that documented the full project cycle from problem identification to reflection. Fourteen portfolios came from primary schools (Grades 5 to 7) and seventeen from secondary schools (Grades 8 to 11). Three researchers independently coded subsets of the data at a manifest level, compared decisions in a shared spreadsheet, resolved disagreements through discussion, and then applied a refined codebook across the full dataset before moving to a deductive, theory-informed interpretation.

Four interconnected themes emerged. The first, systems thinking as a sustainability competency, captured how learners stopped treating community problems as isolated events and began reading them as parts of wider social, institutional and environmental systems. Food insecurity was linked to unemployment, poverty and low academic performance; bullying and substance abuse were traced to family stress, social exclusion and weak support structures. One portfolio observed that mental health issues are both a cause and an effect of poor academic performance — a formulation the authors cite as a clear expression of learners recognising interdependence rather than linear cause-and-effect chains.

The second theme, problem-solving through solution design, revealed a striking pattern: similar problems produced different solutions in different contexts. Littering was met with clean-up campaigns in some schools, recycling schemes or environmental education in others; food insecurity prompted vegetable gardens in one community and donation or sponsorship drives in another. From a variation-theory perspective, this divergence is significant. It suggests learners were not mechanically applying a fixed template but had discerned which aspects of their particular situation mattered most — infrastructure failure, behavioural practices or municipal service delivery — and adjusted their strategies accordingly. Learners also reflected on barriers such as limited funding, low participation and time constraints, and revised their plans in response.

The third theme showed learning happening through real-world civic action itself. Learners investigated problems, spoke to stakeholders, implemented solutions and then reflected on how their own understanding had shifted. The portfolios record increased awareness, emotional maturity, empathy and insight after learners engaged directly with community members affected by poverty, gender-based violence, mental health struggles and environmental degradation. This aligns with the experiential learning tradition running from John Dewey through David Kolb’s cycle of concrete experience, reflective observation, abstract conceptualisation and active experimentation — a cycle the YCAP process, with its toolkit-guided structure and mentor support, enacts in full.

The fourth theme, civic responsibility and learner agency, is perhaps the most striking. Learners described themselves with the language of responsibility, leadership, ownership and hope, positioning themselves as capable actors who could mobilise others rather than waiting for adults or institutions to intervene. Statements such as “We want to be role models for our peers” reflect what the researchers identify as normative thinking: not just judging what is wrong, but making value-based claims about the kind of community they want to help create. The authors connect this to Gert Biesta’s concept of subjectification — education forming autonomous civic subjects rather than merely qualified graduates — and to Westheimer and Kahne’s distinction between participatory and justice-oriented citizenship, the latter targeting the structural roots of social problems rather than their visible symptoms.

Taken together, the four themes describe a progression from recognising problems, to designing responses, to learning through action, to acting as civic subjects. The authors argue that variation theory supplies the explanatory mechanism the sustainability education literature has been missing: the field is strong on describing competency outcomes but weak on explaining how they are produced, particularly at school level, where most evidence comes from higher education and engineering contexts in the Global North. In YCAP, variation arose through the different community challenges teams tackled, through mentors deliberately exposing learners to other teams’ projects, and through the diversity of solutions generated even for similar problems. The study is careful to note that it identifies patterns consistent with variation theory rather than proving causal relationships.

The practical implications are concrete. The authors suggest that mentors should embed exposure to variation deliberately rather than leaving it to chance, that school leaders should structure opportunities for learners to engage a wider range of stakeholders such as local government and health services — engagement that was inconsistently present across the dataset — and that policymakers consider toolkit-based, reflection-oriented models like YCAP as a viable way to extend sustainability education beyond universities. Future programme design, they add, could incorporate the two competencies absent from the data: anticipatory thinking, through activities requiring learners to forecast long-term consequences before implementation, and genuine collaborative competence, through shared decision-making with communities rather than purely informational liaison. The study’s limitations are acknowledged — the sample comprised only recognised, high-quality portfolios, and the analysis was confined to what learners documented — but the central message stands. When a structured process holds the pedagogical architecture constant while community context varies, young learners can discern the critical features of sustainability challenges in ways that decontextualised classroom instruction cannot produce. In this model, the community is the curriculum, variation is the mechanism, and the formation of autonomous civic subjects is the goal.

Subject of Research: Sustainability competencies and civic learning in school-based community projects analysed through variation theory

Article Title: Exploring civic learning and sustainability competencies in school community projects using variation theory

Article References: Osman, A., Sigenu, A., & du Plessis, C. (2026). Exploring civic learning and sustainability competencies in school community projects using variation theory. Discover Education, 5(1), Article 1131. https://doi.org/10.1007/s44217-026-02245-8

Image Credits: AI Generated

DOI: 10.1007/s44217-026-02245-8

Keywords: sustainability education, civic learning, variation theory, systems thinking, normative thinking, strategic thinking, experiential learning, Youth Citizen Action Programme, South Africa, school community projects, learner agency, qualitative research

The timing of the proposal is significant. Higher education systems worldwide are straining under rising application volumes, constrained resources, and intensifying expectations around widening participation. In the United Kingdom alone, the 2023 UCAS cycle involved more than 750,000 applicants and nearly three million applications. Admissions systems designed for smaller applicant pools and standardised qualifications are buckling under the load, and AI promises scalability. Yet Alenezi’s review of the literature reveals a troubling pattern: most AI-enabled admissions scholarship evaluates systems purely on prediction accuracy, enrolment optimisation, or processing efficiency, while paying scant attention to learner agency, transparency, or educational equity.

The risks are not hypothetical. The paper documents how algorithmic systems trained on historical data can encode and perpetuate existing inequalities. Amazon’s experimental CV-screening tool systematically downgraded applications from women because it had learned from historically male-dominated hiring patterns. The COMPAS recidivism tool demonstrated racially disparate predictions. And the UK’s 2020 A-level algorithm depressed grades for students from lower-performing schools, sparking a national controversy. In admissions, the danger takes specific technical forms: representation bias when training datasets under-represent particular groups, measurement bias when proxy variables like postcode or school type capture group membership rather than merit, and feedback bias when predictions shape institutional actions that then create self-fulfilling cycles.

HCPAF responds with an architecture of five interconnected dimensions: learner profiling and contextual understanding; explainable AI decision support; equity and inclusion auditing; human oversight and shared decision-making; and personalised transition and success support. Crucially, the framework treats these dimensions as recursive rather than sequential. Profiling choices shape what can be explained and audited; fairness findings may force revisions to model features; human review can trigger model correction; and transition outcomes feed back into assumptions made at the point of admission. The result is a system of mutual constraints rather than a frictionless pipeline.

Technically, the framework leans on established explainable AI techniques such as LIME and SHAP, which generate interpretable accounts of how input variables contribute to model outputs. But Alenezi, drawing on the work of Khosravi and colleagues, insists that explainability must be calibrated to its audience. An explanation intelligible to a data scientist may be opaque to an admissions officer or an applicant, so the framework calls for layered explanation architectures serving multiple stakeholder groups simultaneously. All AI-generated recommendations must be accompanied by plain-language explanations, consistent with GDPR Article 22 rights and the EU AI Act’s requirements for high-risk systems in education.

The framework is equally candid about trade-offs that much of the AI-in-education literature glosses over. Richer contextual data can support fairer decisions, but collecting more socioeconomic, demographic, and behavioural information increases privacy risk. Postcode-level measures such as POLAR4 quintiles can identify structural disadvantage, yet the same variables may act as proxies for ethnicity, class, or disability — and removing them entirely can make disadvantage statistically invisible. HCPAF’s answer is neither uncritical inclusion nor blanket exclusion, but counterfactual and subgroup sensitivity testing, with contextual variables used preferentially to expand opportunity rather than deny access. Similarly, the framework acknowledges that demographic parity, equalised odds, and individual fairness cannot always be satisfied simultaneously, and that carrying admissions data into student-support systems may enable earlier help while creating risks of labelling and surveillance.

One of the study’s most distinctive conceptual contributions is its treatment of trust. Rather than assuming that transparency automatically builds confidence in AI systems, Alenezi frames trust as a relational, calibrated, and empirically contingent construct. More information does not necessarily generate more trust: disclosure of uncertainty, data limitations, or unequal error rates may appropriately reduce confidence. The relevant objective is calibrated trust, in which applicants and staff understand what the AI can and cannot do, can challenge its outputs, and have credible evidence that errors and inequities will be detected and remedied. Explainability, fairness auditing, and human oversight may reinforce one another — or they may reveal justified grounds for distrust.

Generative AI occupies a carefully bounded role within the framework. While predictive machine learning supports profiling and fairness monitoring, large language models are positioned mainly as an interaction and explanation layer: translating model outputs into applicant-accessible language, providing source-grounded conversational guidance about programmes and procedures, and personalising pre-arrival information. Recognising that LLMs can hallucinate, reproduce bias, and generate confident but incorrect advice, HCPAF requires retrieval from authoritative institutional sources, logging and audit of outputs, clear disclosure that applicants are interacting with AI, and escalation to human advisers wherever financial, disability, visa, safeguarding, or admissions-condition issues arise. Notably, the framework treats AI-assisted personal statements as a governance problem rather than an automatic integrity violation, urging institutions to avoid unreliable AI-detection tools and to ensure applicants are not disadvantaged by unequal access to language models.

The theoretical scaffolding is deliberately broad, integrating six traditions: Shneiderman’s human-centred AI, responsible AI and algorithmic governance, Sen’s Capability Approach, educational equity theory, Value-Sensitive Design, and human–AI collaboration theory. The Capability Approach in particular reframes the evaluation question: admissions systems should be judged not only by whether they predict academic success accurately, but by whether they expand or constrain the substantive freedoms of diverse learner populations to access and flourish within higher education. The framework also identifies normative alignments with Sustainable Development Goals 4, 10, and 16, though Alenezi is careful to stress that these are alignments of principle, not evidence of impact.

The author is refreshingly explicit about the framework’s limits. HCPAF is a conceptual synthesis, not an empirically validated admissions model, and prospective evidence on live AI-assisted admissions remains thin. The paper’s illustrative scenarios — a post-92 university using contextualised offer-making with monthly equity audits, a research-intensive university deploying a generative AI advising portal with human review of consequential guidance, and a specialist provider connecting admissions profiles to early-warning retention systems — are explicitly hypothetical design propositions rather than documented deployments. Until validation evidence accumulates through case studies, participatory co-design, and longitudinal evaluation, HCPAF should be used as a governance heuristic for interrogating AI-enabled admissions, not as proof that such systems will improve equity or trust. That honesty may prove to be the framework’s most important feature: at a moment when institutions face intense pressure to adopt AI, it offers a roadmap for doing so on educational rather than purely institutional terms.

Subject of Research: A human-centred AI and learning analytics governance framework for equitable personalised higher education admissions

Article Title: A human-centred artificial intelligence and learning analytics framework for equitable personalised higher education enrolment

Article References: Alenezi, A. F. (2026). A human-centred artificial intelligence and learning analytics framework for equitable personalised higher education enrolment. Discover Education, 5(1), Article 1084. https://doi.org/10.1007/s44217-026-02221-2

Image Credits: AI Generated

DOI: 10.1007/s44217-026-02221-2

Keywords: human-centred AI, learning analytics, higher education admissions, educational equity, explainable AI, generative AI, learner agency, algorithmic bias, responsible AI governance, personalised learning, widening participation, student success

At the heart of the framework lies a three-level distinction that challenges how VR research is often reported. Technological immersion refers to objectively describable properties of a media system, such as representational fidelity, interactivity, sensorimotor contingency, and the range of sensory channels addressed. Subjective immersive experience, by contrast, describes the temporally variable ways learners perceive and interpret their interaction with a virtual environment, including presence, engagement, embodiment, and perceived agency. Immersive learning, the third level, denotes a pedagogically designed process in which a potentially immersive activity is connected with prior experiences, preparation, situated tasks, educator guidance, and subsequent reflection and transfer. Confusing these levels, the paper argues, leads to the widespread but unjustified assumption that wearing a headset automatically produces presence, engagement, or understanding.

The theoretical backbone of the model draws on John Dewey’s pragmatist conception of experience, particularly his principles of continuity and interaction. Continuity means that every experience draws on preceding experiences and shapes how later ones are interpreted. Interaction refers to the relationship between a learner’s existing dispositions and the objective conditions of a situation. Under this view, learners never enter a virtual environment as neutral recipients of stimuli; they arrive with prior knowledge, expectations, values, emotions, media experience, and culturally shaped interpretive frameworks. The same virtual scenario may become a challenging learning opportunity for one student, confirm an existing belief for another, or remain confusing or irrelevant for a third. Dewey’s warning that not every experience is educative, and that some may even restrict future learning, sits uncomfortably beside the enthusiasm that often surrounds educational VR.

The I-MEIL organizes VR-supported learning into three interrelated phases: prior and preparatory experience, situated immersive experience, and reflective and transfer-oriented experience. The first phase encompasses learners’ biographical and educational backgrounds alongside instructional preparation, such as activating prior knowledge, clarifying tasks, establishing expectations, and developing technical familiarity. The second phase concerns situated interaction with the designed environment and the cognitive, affective, and excitative response states that may arise within it. The third phase comprises pedagogically supported opportunities to articulate, contextualize, critically appraise, consolidate, and transfer what was experienced. Crucially, these phases are not conceived as automatically completed stages or a fixed causal sequence; they identify analytically distinct yet interdependent parts of a learning design whose relationships vary with learners, tasks, content, facilitation, and context.

Technological features, in this framework, function as affordances rather than direct causes. Drawing on James Gibson’s ecological psychology, the model treats possibilities for action as arising in the relationship between an environment and an acting individual. Manipulating a virtual object may support procedural exploration when aligned with a learning goal, but it may be irrelevant or distracting when peripheral to the task. Multisensory information may aid orientation in one application while imposing unnecessary processing demands in another. Fidelity itself should not be equated with visual realism; a highly realistic representation may be unnecessary or even distracting when schematic information better supports the intended task, while detailed spatial representation may be essential when learners must distinguish subtle features or navigate complex environments.

During the situated phase, the model draws on the Differential Susceptibility to Media Effects Model to describe three transient response states. Cognitive responses concern attention, interpretation, and mental effort; affective responses include interest, curiosity, enjoyment, discomfort, or frustration; excitative responses refer to physiological arousal. These states may fluctuate constantly during a VR activity because of distraction, task difficulty, technical disruption, cybersickness, or deliberate disengagement. Constructs frequently measured in VR studies, such as presence, agency, embodiment, and cognitive load, serve as indicators of selected aspects of this experience, but the model insists they are not interchangeable and not evidence of learning in themselves. A high presence score documents a subjective response; it does not demonstrate conceptual understanding, reflection, or transfer.

The framework also takes a cautious stance on one of VR’s most hyped applications: perspective-taking. Virtual environments can present situations from a particular visual position and invite learners to consider other people’s experiences, but occupying a first-person perspective or controlling an avatar does not reproduce another person’s cognitive standpoint, identity, or lived experience. Learners continue to interpret representations from their own social, cultural, and biographical positions. When simulations address marginalization, discrimination, homelessness, or migration, the model warns against equating perspective-taking with becoming another person, a critique that echoes longstanding concerns about so-called cybernetic tourism and the reduction of situated identities to temporarily consumable representations.

Reflection and transfer form the third instructional function, and the model argues they cannot be assumed to occur spontaneously. Learners may remove the headset with fragmented impressions, unresolved emotions, misconceptions, or unanswered questions. Post-VR reflection has distinctive objects: learners can examine not only the represented content but also how technological mediation shaped what they noticed, felt, or were able to do, and what the simulation omitted or whose perspective it privileged. Generative activities such as self-explanation, summarizing, and collaborative discussion can support integration, but the paper cites evidence that strategies effective in other contexts may yield no additional benefit in procedural VR learning and may even introduce competing demands. Preparation and reflection, in other words, are conditional design elements rather than universally effective additions.

The model’s implications for research are equally demanding. Studies should distinguish technological conditions, subjective response states, instructional processes, and learning outcomes, combining self-reports, behavioral and interaction data, physiological measures, and outcome assessments rather than relying on any single indicator. Longitudinal and process-oriented designs are prioritized because the educational meaning of a VR activity may emerge across time rather than during the virtual episode alone. For designers, the framework offers a sequence of questions rather than a prescription: identify a genuine educational or biographical need, compare VR against alternatives such as video, desktop simulation, or role-play, and select the technology only when its representational and interactive conditions contribute meaningfully to the task. Learning outcomes themselves are treated as multidimensional, spanning cognitive, affective, attitudinal, behavioral, and psychomotor domains, and as recursive: new understanding may generate fresh questions and needs, reconnecting the virtual episode with the learner’s continuing stream of experiences. In an era of accelerating VR adoption, the I-MEIL offers a sobering but constructive message: the headset is where the technology begins, not where the learning ends.

Subject of Research: A conceptual model of how virtual reality-supported immersive learning connects prior experience, preparation, situated engagement, reflection, and transfer

Article Title: An integrative model of virtual reality supported immersive learning within a biographical stream of experiences

Article References: Mulders, M. (2026). An integrative model of virtual reality supported immersive learning within a biographical stream of experiences. Discover Education, 5(1), Article 1070. https://doi.org/10.1007/s44217-026-02219-w

Image Credits: AI Generated

DOI: 10.1007/s44217-026-02219-w

Keywords: virtual reality, immersive learning, educational technology, instructional design, experiential learning, presence, embodiment, cognitive load, reflection, transfer of learning, Dewey, learner agency