The researchers drew on the Unified District Information System for Education Plus, the Indian government’s exhaustive annual school census, for the academic years 2020–21 and 2021–22. From these records they assembled a portrait of each district’s educational machinery: ten infrastructure variables, including separate rooms for headmasters, functional electricity, playgrounds, libraries, furniture, toilets, drinking water, and computer facilities, were compressed into a single composite index using principal component analysis. Against this resource base they measured outputs such as promotion rates, girls’ enrollment shares, the number of schools, and the percentage of trained teachers.

To judge which districts squeeze the most schooling from their inputs, the team deployed Data Envelopment Analysis, a non-parametric linear programming technique pioneered by Charnes, Cooper, and Rhodes in 1978. DEA works by constructing an efficiency frontier from the best-performing districts and then scoring every other district by its distance from that frontier, with a score of 1 denoting full technical efficiency. The analysis was run under both constant and variable returns to scale assumptions, using an input-oriented model that asks how much a district could proportionally trim its resources while holding outcomes fixed, a reasonable framing since administrators control budgets more tightly than results.

The headline finding is that Kargil and Pulwama sit on the frontier across both years and all three education levels, primary, upper primary, and secondary, in both CRS and VRS models. These two districts, one in Ladakh’s high-altitude desert and one in the Kashmir Valley, achieved steady improvements in outcomes as teacher training expanded while other inputs changed little. Ganderbal and Rajouri proved efficient in primary and upper primary education, while Kathua, Leh, and Srinagar demonstrated efficiency in primary and secondary schooling. Mean VRS efficiency scores were remarkably high throughout: 0.980 for primary schools in 2020–21, 0.952 for upper primary, and 0.943 for secondary, with only marginal drift by the following year.

Beneath the averages, the returns-to-scale diagnostics reveal where resources are being misallocated. In 2020–21, primary schools in Baramulla and Jammu exhibited decreasing returns to scale, meaning a 1 percent increase in all inputs yielded less than 1 percent more output, a signature of input congestion. Both recovered the next year as infrastructure and trained-teacher percentages improved. By 2021–22, every district’s primary schools showed either increasing or constant returns to scale. Doda and Reasi, though below average efficiency, displayed increasing returns to scale, suggesting these underperforming districts would benefit more from additional resources than from austerity, since their schools have absorptive capacity waiting to be filled.

Efficiency alone, however, is a static snapshot. To capture whether districts are getting better or worse over time, the study applied the Malmquist Productivity Index, which measures total factor productivity change between two periods as the geometric mean of distance functions relative to shifting frontiers. The index decomposes into technical efficiency change, itself split into pure and scale components, and technological change, which tracks movement of the frontier itself. An index above 1 signals improvement; below 1, decline.

For primary schools, the verdict was sobering: the average Malmquist index was 0.905, a 9.5 percent productivity decline, driven almost entirely by technological change of 0.906. Nine districts recorded productivity losses, with the steepest in Samba at 78.2 percent, Rajouri at 51.7 percent, Kishtwar at 47.7 percent, and Kathua at 34 percent. Samba’s technological component alone fell by roughly 77.2 percent alongside declines in technical and pure efficiency. Yet thirteen districts, including Anantnag, Badgam, Bandipora, Baramulla, Doda, Ganderbal, Kargil, Kulgam, Kupwara, Pulwama, Ramban, Reasi, and Shopian, posted productivity gains, some attributable purely to technological advancement and others to improvements in management and scale efficiency.

The most consequential pattern emerges at the top of the system. In upper primary schools, average technical, pure, scale, and total factor productivity indexes all equaled 1, with only technological change declining to 0.894; Kargil’s productivity surged 263.6 percent through technological change. But in secondary schools, the picture darkens: productivity ranged from 0.463 to 1.000, fifteen districts recorded declining total factor productivity, and not a single secondary school district demonstrated an increase. Seven districts, including Bandipora, Doda, Ganderbal, Kupwara, Punch, Reasi, and Shopian, held perfectly steady. The study’s authors note that while primary and upper primary units show productivity growth from technical, technological, pure, and scale factors, secondary schooling’s changes flow solely through technological channels, with no efficiency gains to offset frontier slippage.

The two Ladakh districts, Leh and Kargil, performed well across all three levels, which the researchers attribute to central government funding of their schools. The policy prescriptions are region-specific: Ladakh needs primary schools upgraded to upper primary or secondary status to improve retention, plus infrastructure hardening for harsh weather, transportation, and in-service teacher training. In Jammu and Kashmir, trained teachers must be deployed to rural areas, with new schools in Kupwara, Baramulla, Rajouri, and Poonch, supported by residential facilities, hardship allowances, and safety measures.

The study’s limitations are candid: a two-year window, a restricted set of input and output variables, no district-level learning outcomes, and no second-stage regression analysis. Still, the core message resonates far beyond the Himalayas. Increased educational expenditure and more teachers do not reliably translate into better schooling; several underperforming districts show high student-teacher ratios and underutilized infrastructure, hallmarks of resource misallocation rather than input shortage. Optimizing existing resources, the authors conclude, is more likely to lift productivity than indiscriminate expansion, a finding with direct implications for achieving Sustainable Development Goal 4 in India’s newest Union Territories and beyond.

,

Subject of Research: A district-level analysis of school efficiency and productivity in Ladakh and Jammu & Kashmir finds Kargil and Pulwama consistently on top, while secondary schools show no productivity gains.

Article Title: Evaluating the efficiency and productivity change of primary and secondary school education in Ladakh and Jammu & Kashmir

Article References: Hussain, A., & Rizvi, H. S. (2026). Evaluating the efficiency and productivity change of primary and secondary school education in Ladakh and Jammu & Kashmir. Discover Education, 5(1), Article 834. https://doi.org/10.1007/s44217-026-02100-w

Image Credits: AI Generated

DOI: 10.1007/s44217-026-02100-w

Keywords: school efficiency, Data Envelopment Analysis, Malmquist index, Ladakh, Jammu and Kashmir, UDISE+, education productivity, SDG 4, teacher training, Kargil, Evaluating, efficiency

The study, published in SN Social Sciences, examined the state of education in Gosaba through secondary data from India’s official education and census databases. Researchers found that student enrolment dropped from 41,661 in the 2012-13 academic year to just 30,922 by 2016-17, a decline of more than 10,000 students. During the same period, the dropout rate surged from 4.87 percent to 16.78 percent, signaling a deepening crisis in educational retention across the deltaic landscape.

Gosaba Community Development Block sits within the Indian Sundarban Delta, a UNESCO World Heritage region characterised by its mangrove forests, tidal waterways, and extreme vulnerability to cyclones and sea-level rise. The geographic isolation of the area, combined with recurrent natural disasters, creates unique barriers to educational delivery that are rarely captured in national-level assessments. The researchers argue that this micro-level evidence is essential for understanding how Sustainable Development Goal 4, which calls for inclusive and equitable quality education, can be achieved in ecologically fragile zones.

The analysis drew on data from the Unified District Information System for Education, the Census of India 2011, and the National Institute of Educational Planning and Administration. Using descriptive statistics and Pearson correlation analysis, the team evaluated school distribution, enrolment patterns, dropout rates, teacher characteristics, pupil-teacher ratios, and school infrastructure across the block. The methodological approach allowed the researchers to identify relationships between physical infrastructure, teacher availability, and learning outcomes in a region where such connections had previously gone undocumented at this scale.

Infrastructure findings revealed stark disparities that persist despite some improvements over the study period. Only 27 percent of schools in Gosaba had access to electricity, while a mere 8 percent offered computer facilities to students. These figures stand in sharp contrast to national averages and underscore the digital divide that separates deltaic communities from urban educational centres. The correlation analysis demonstrated that schools with better infrastructure tended to have better teacher availability and stronger learning outcomes, suggesting that physical resources and human capital are deeply intertwined in shaping educational quality.

The pupil-teacher ratio emerged as a critical variable in the study’s findings. In regions where teacher shortages compound infrastructural deficits, the quality of instruction suffers disproportionately. The researchers noted that schools with favourable ratios correlated positively with learning outcomes, reinforcing established evidence that teacher availability is one of the strongest predictors of educational quality in developing contexts. The findings align with broader scholarship on the knowledge capital of nations, which demonstrates that cognitive skills and human capital formation are fundamental drivers of long-term economic growth.

The environmental context of the Sundarban Delta cannot be separated from its educational challenges. The region faces escalating threats from climate change, including intensified cyclonic activity, salinisation of agricultural land, and gradual submergence of low-lying areas. These environmental stressors disrupt schooling through physical damage to educational facilities, displacement of communities, and economic pressures that force families to prioritise immediate livelihood needs over education. The Intergovernmental Panel on Climate Change has identified such regions as particularly vulnerable to the intersection of environmental degradation and socio-economic marginalisation.

The study’s authors argue that conventional, one-size-fits-all educational policies are inadequate for regions like the Sundarban Delta. Instead, they call for context-specific and integrated approaches that simultaneously address educational infrastructure, teacher deployment, and environmental resilience. Such policies would need to account for the unique geographical constraints of deltaic communities, where access to schools may be limited by waterways, where cyclone preparedness must be built into educational planning, and where livelihood insecurity directly affects school attendance and retention.

The research contributes to a growing body of evidence highlighting the challenges of achieving Sustainable Development Goal 4 in marginalised regions. Previous studies across South Asia have documented similar patterns, where infrastructure deficits, teacher shortages, and socio-economic vulnerabilities combine to produce poor educational outcomes. The Gosaba findings add a crucial environmental dimension to this literature, demonstrating that ecological fragility is not merely a backdrop to educational inequality but an active force that shapes access, quality, and continuity of learning.

As the international community approaches the 2030 deadline for the Sustainable Development Goals, the Gosaba study serves as a stark reminder that progress toward quality education remains uneven and that the most vulnerable communities continue to be left behind. The researchers emphasise that without targeted interventions addressing both educational and environmental challenges in regions like the Indian Sundarban, the global promise of inclusive, equitable quality education for all will remain unfulfilled for millions of children living at the frontlines of climate change.

Subject of Research: Quality education and SDG 4 achievement in the Gosaba Community Development Block of the Indian Sundarban Delta

Article Title: A progressive study on quality education (SDG 4) in Gosaba Community Development Block of the Indian Sundarban Delta

Article References: Pal, A., Hansda, T., & Ghosh, T. (2026). A progressive study on quality education (SDG 4) in Gosaba Community Development Block of the Indian Sundarban Delta. SN Social Sciences, 6(9), Article 365. https://doi.org/10.1007/s43545-026-01650-y

Image Credits: AI Generated

DOI: 10.1007/s43545-026-01650-y

Keywords: SDG 4, quality education, Sundarban Delta, Gosaba, school infrastructure, dropout rate, pupil-teacher ratio, climate vulnerability, educational policy, India, rural education, sustainable development

The study, led by Oladejo Olufemi Michael of the University of KwaZulu-Natal, together with colleagues from Rhodes University and York St. John University, set out to investigate how and to what extent stakeholder support influences faculty performance within resource-limited settings. Grounded in Stakeholder Theory and Social Exchange Theory, the research tested five hypotheses using a quantitative design. Data were collected from 197 academic staff across three public universities in South-West Nigeria: the University of Ibadan, Lagos State University, and Federal University Oye-Ekiti. These institutions represent three distinct generations of Nigerian higher education, providing a broad view of the academic experience across diverse institutional histories.

The researchers employed stratified random sampling to distribute 256 questionnaires, of which 202 were returned, yielding a response rate of 78.9 per cent. After excluding five partially completed questionnaires, 197 fully usable responses remained. The sample size proved adequate for the analytical technique employed: Partial Least Squares Structural Equation Modelling, conducted using SmartPLS 4. Applying the 10-times rule recommended for this method, a minimum of 40 cases was required given the maximum of four predictors pointing to any single endogenous construct. The achieved sample exceeded this threshold by approximately five times.

The measurement instrument comprised four latent constructs assessed on a five-point Likert scale. Key Stakeholders’ Support was measured with seven items and demonstrated a Cronbach’s alpha of 0.934. Training and Development Plans comprised four items with an alpha of 0.813. Training and Development Programme included five items with an alpha of 0.881. Academic Staff Performance was captured through six items measuring teaching quality, research output, and community service, with an alpha of 0.880. All constructs exceeded recommended reliability thresholds, and discriminant validity was confirmed using the Heterotrait-Monotrait ratio, with all values falling below the 0.85 threshold.

The results revealed a significant direct effect of stakeholder support on academic staff performance, with a path coefficient of 0.423, a t-value of 5.935, and a p-value below 0.001. This finding aligns with Social Exchange Theory, which holds that social behaviour is governed by the norm of reciprocity. When academic staff perceive their institutional environment as genuinely supportive, they reciprocate through greater performance effort. The f-squared value of 0.215 represents a medium effect size, indicating that the overall quality of the stakeholder support environment is itself a meaningful organisational resource for academic staff, independent of any training infrastructure it may enable.

Perhaps the most structurally important finding concerns the relationship between training and development plans and training and development programmes. The path coefficient between these two constructs was 0.709, with a t-value of 16.272 and a p-value below 0.001. The f-squared value of 1.012 indicates a large effect size. This path explained 50.3 per cent of the variance in training and development programmes. The researchers interpret this as evidence that once rigorous planning is properly funded, translation into action follows smoothly. The primary bottleneck, they argue, is not the conversion of plans into active programmes but the failure of institutions to produce adequately resourced plans in the first place.

The study’s fourth hypothesis confirmed that training and development programmes fully mediate the relationship between training and development plans and academic staff performance. The indirect effect was significant at 0.210, with a 95 per cent confidence interval of 0.090 to 0.353 that excludes zero. Critically, the direct path from training and development plans to academic staff performance was not significant, indicating that plans alone produce no measurable performance benefit. Full mediation through programme execution is the decisive pathway. Together, stakeholder support and training and development programmes explained 56.2 per cent of the variance in academic staff performance.

A further exploratory finding emerged from the fifth hypothesis, which tested a feedback path from programme execution back to stakeholder support. The path coefficient was 0.511, with a t-value of 5.662 and a p-value below 0.001. This finding extends Freeman’s instrumental framework and adds nuance to Mitchell, Agle, and Wood’s salience theory by suggesting that visible programme quality can reinforce stakeholder commitment rather than being solely a precondition for programme delivery. The researchers argue that when universities successfully translate plans into programmes, the visible quality of those programmes strengthens the confidence of key stakeholders, who in turn intensify their support. This virtuous cycle has not previously been documented in the African higher education literature.

Data collection observations suggest that government and university administration, despite their formal authority, are among the least effective providers of support in practice. Students, by contrast, contributed meaningfully through informal encouragement and advocacy. The researchers note that these observations are descriptive and exploratory, as the study measured stakeholder support as a composite construct and cannot produce differential structural path estimates by stakeholder type. They recommend that future studies disaggregate this construct into separate sub-constructs for government, university administration, private sector, and students to formally test whether differential effects exist.

The study carries several practical implications for university leaders and policymakers working to advance SDG 4. The researchers argue that policy documents and strategic plans are necessary but insufficient. Decisive progress depends on funded, executed, and monitored training programmes backed by coordinated multi-stakeholder commitment. They recommend that government agencies and institutional funders secure adequate budgets before rolling out new programmes, that university administrators foster supportive climates through transparent resource allocation and meaningful recognition of academic labour, and that institutions establish coordination platforms to align the contributions of all four stakeholder groups around shared professional development goals integrated with SDG 4 monitoring frameworks.

Subject of Research: Higher Education Research

Article Title: Future proofing the academy: the nexus between stakeholder support and academic staff performance

Article References: Olufemi Michael, O., Sybert, M., Nene Ernest, K., & Olayiwola Idris, G. (2026). Future proofing the academy: the nexus between stakeholder support and academic staff performance. Frontiers in Education, 11, Article 1857809. https://doi.org/10.3389/feduc.2026.1857809

Image Credits: AI Generated

DOI: 10.3389/feduc.2026.1857809

Keywords: academic staff performance, stakeholder support, SDG 4, Nigerian public universities, training and development, PLS-SEM, Stakeholder Theory, Social Exchange Theory, higher education, sub-Saharan Africa, Future, proofing

The study was led by Omar Abdi Mohamud and Mohamed Jama Mohamed of the Graduate School at SIMAD University in Mogadishu, together with Leila Mohamed Warsame and Abdirahman Ibrahim Abdi of the university’s Faculty of Education. Their work addresses a persistent gap in the literature: while access to secondary education has expanded substantially in Somalia in recent years, empirical evidence about what actually drives quality in Somali classrooms has remained scarce. Education quality, the authors note, remains a critical challenge in developing and post-conflict countries, and Somalia is a particularly striking case, having spent decades reconstructing institutions that were shattered by civil war.

To untangle the relationships between potential determinants and educational outcomes, the team employed a quantitative cross-sectional survey design. They gathered data from 239 respondents drawn from four distinct groups: students, teachers, parents, and school administrators. Participants were selected through simple random sampling within the participating schools, a technique that gives every member of each group an equal chance of inclusion and helps reduce selection bias in a setting where official records can be incomplete. The survey instrument captured perceptions and assessments across the four hypothesized determinants of quality—teacher competence, favorable student–teacher conditions, curriculum relevance, and parental involvement—alongside measures of overall education quality.

The analytical engine of the study was Partial Least Squares Structural Equation Modeling, or PLS-SEM, conducted with the software SmartPLS 4. Structural equation modeling is a family of statistical techniques that allows researchers to test networks of relationships among observed measurements and theoretical constructs simultaneously, rather than examining each variable in isolation. PLS-SEM, in particular, is favored in social science research when the goal is to explain variance in a target outcome and when constructs are measured through multiple indicators. Before testing their hypotheses, the researchers verified that the measurement model demonstrated satisfactory reliability and validity, meaning the survey items consistently and accurately captured the concepts they were intended to represent.

The results of the structural model were clear and, in one respect, surprising. Curriculum relevance emerged as the strongest predictor, with a path coefficient of β = 0.303 and a p-value below 0.001, indicating a highly significant positive association with education quality. Favorable student–teacher conditions followed closely at β = 0.273, also with p < 0.001. Parental involvement showed a significant association at β = 0.213 with p = 0.015, and teacher competence, while still significant, was the weakest of the four at β = 0.122 with p = 0.038. Together, these four hypotheses—labeled H1 through H4 in the study—were all supported by the data.

Perhaps the most striking finding for anyone who assumes that teachers are the whole story is the relative ranking. Teacher competence mattered, but it was the smallest contributor in the model. What mattered more was whether the curriculum felt relevant to students’ lives and futures, and whether the conditions surrounding the student–teacher relationship—classroom climate, mutual respect, and the day-to-day dynamics of learning—were favorable. This pattern suggests that in Mogadishu’s secondary schools, the surrounding architecture of education, from syllabus design to classroom atmosphere to family engagement, may matter as much as the individual skill of the person at the front of the room.

The model as a whole explained 58.5 percent of the variance in education quality, a statistic the researchers describe as indicating moderate explanatory power. In practical terms, this means that more than half of the differences in perceived education quality across the surveyed schools could be accounted for by the combination of these four factors, while the remainder reflects other influences not captured in the model—potentially including infrastructure, resources, leadership practices, and community conditions. For a single study in a complex, post-conflict environment, an R² of 0.585 represents a substantial and meaningful share of explained variation.

The authors situate their findings within two established theoretical frameworks: Education Production Function Theory and Systems Theory. Education Production Function Theory treats schooling as a process in which inputs—teachers, curricula, family resources—are transformed into educational outputs, allowing researchers to estimate which inputs yield the greatest returns. Systems Theory, meanwhile, emphasizes that the components of an organization are interconnected, so that a change in one part reverberates through the whole. The Mogadishu data support both views: education quality was associated with interconnected instructional, classroom, curricular, and family-related conditions rather than with any single lever pulled in isolation.

The implications reach into international development policy. The study explicitly connects its findings to Sustainable Development Goal 4, the United Nations commitment to inclusive and equitable quality education and lifelong learning opportunities for all. For countries emerging from conflict, SDG 4 has often been pursued primarily through enrollment drives—getting children through the school door. This research suggests that the harder question, what happens once students are inside, depends on strengthening curriculum implementation, investing in teacher professional development, cultivating favorable student–teacher conditions, and building genuine school–parent partnerships. The authors argue that these steps can support Somalia’s progress toward a more resilient and sustainable education system.

The research also carries a methodological significance. Conducting rigorous survey research in Mogadishu requires navigating security concerns, institutional fragility, and the practical difficulties of sampling in a rapidly changing urban environment. The study received ethical approval from the SIMAD University Centre for Research and Development Institutional Research Ethics Committee, was conducted in accordance with the Declaration of Helsinki, and obtained written informed consent from all participants, who took part voluntarily and could withdraw at any time. The work was supported by the Center for Research and Development at SIMAD University, a homegrown institution producing context-specific evidence from within Somalia rather than about it. As the country continues its long reconstruction, studies like this one provide the statistical footing on which education policy can finally stand—evidence that in the classrooms of Mogadishu, relevance, relationships, families, and teachers together shape whether quality education takes root.

Subject of Research: Determinants of secondary education quality in Mogadishu, Somalia, analyzed with partial least squares structural equation modeling

Article Title: Determinants of education quality in secondary schools in Mogadishu Somalia using partial least squares structural equation modeling

Article References: Mohamud, O. A., Warsame, L. M., Mohamed, M. J., & Abdi, A. I. (2026). Determinants of education quality in secondary schools in Mogadishu Somalia using partial least squares structural equation modeling. Discover Sustainability. https://doi.org/10.1007/s43621-026-04740-2

Image Credits: AI Generated

DOI: 10.1007/s43621-026-04740-2

Keywords: education quality, secondary education, Somalia, Mogadishu, PLS-SEM, teacher competence, curriculum relevance, parental involvement, student–teacher conditions, SDG 4, post-conflict education, Discover Sustainability

The research, led by Sathya M. and Alamelu R. of SASTRA Deemed to Be University in India, responds to a problem that has long haunted education systems worldwide. A UNESCO global monitoring report cited in the study found that fewer than 40 percent of teachers feel adequately prepared to apply digital pedagogy in their classrooms. Most existing professional development programs, according to systematic reviews of the field, are generic, overly theoretical and disconnected from the realities of teaching, with no mechanisms for personalisation, scalability or iterative feedback. The consequence is poor engagement and skills that rarely transfer into day-to-day classroom practice.

SmartPD anchors its ambitions in DigCompEdu, the European Framework for the Digital Competence of Educators, which maps educator skills across six areas including digital resources, teaching and learning, assessment, empowering learners and facilitating learners’ digital competence. Rather than treating the framework as a static checklist, the researchers operationalised it as the backbone of an intervention, aligning each training module with specific DigCompEdu domains and measuring growth with the framework’s Check-In self-assessment tool, which showed acceptable to good internal consistency across all domains, with Cronbach’s alpha values ranging from 0.744 to 0.852.

Methodologically, the study is built on Design-Based Research, an iterative approach in which an intervention is continuously refined through cycles of design, enactment and evaluation in real educational settings. The team moved through four phases: a contextual exploration of teachers’ digital competence, the design and development of SmartPD modules, a short-term implementation with 32 in-service teachers drawn from multiple disciplines, and an evaluation phase using analytics and participant feedback. Even within the two-week pilot, the researchers treated daily prompt adjustments and refinements of AI-generated feedback as micro-iterations, keeping the design responsive rather than fixed.

The technology stack is where the model earns its name. Interactive micro-lessons were built with H5P and organised into four modules covering digital resource creation, learner engagement, digital assessment and digital collaboration. Flipgrid hosted asynchronous video reflections, allowing teachers to review and critique one another’s digital practices. ChatGPT provided immediate, formative feedback on open responses and quiz tasks, offering hints and performance summaries that encouraged self-regulated learning. Underneath it all, an xAPI-based analytics dashboard tracked time on task, completion rates, discussion participation and revision behaviour, visualising engagement data for both facilitators and participants in real time.

The statistical outcomes were unusually emphatic. Paired-sample t-tests revealed significant pre-to-post gains in all five measured DigCompEdu domains, with p-values below 0.001. Effect sizes, calculated as Cohen’s d using the standard deviation of difference scores, reached 5.963 for facilitating learners’ digital competence and 5.565 for teaching and learning, followed by assessment at 4.700, digital resources at 4.640 and empowering learners at 3.344. These figures dwarf the moderate-to-large effects, typically between 0.65 and 0.95, reported in comparable professional development studies, and the authors attribute the magnitude to the synergy between micro-learning granularity and AI-mediated feedback loops.

Correlation analysis added a nuanced picture of how competencies developed. Empowering learners and facilitating learners’ digital competence showed a moderate, statistically significant relationship, with a Pearson correlation of 0.491 and a p-value of 0.004, suggesting that teachers who grew more confident in supporting their students also became better at guiding students’ own digital skill growth. Most other domain pairings were weak or non-significant, which the researchers interpret as a feature of the modular design: because each competency was addressed by dedicated micro-learning units, teachers built discrete skill areas independently and at their own pace rather than through cross-domain tasks.

Engagement and satisfaction data reinforced the quantitative story. xAPI analytics recorded high module completion rates and strong participation in micro-learning activities, while qualitative data from reflective journals, focus groups and AI feedback logs surfaced three dominant themes: perceived usefulness of AI-generated feedback, the flexibility and relevance of the micro-learning format, and the value of collaborative reflection. Teachers described the AI feedback as timely and specific, credited it with identifying areas for improvement, and reported greater confidence in selecting and deploying digital tools in their own lesson planning.

The authors are careful to frame the limits of the evidence. The sample of 32 participants is small, the evaluation window was short, and self-reported measures carry the risk of response bias. The reliance on reliable internet access and adequate equipment could exclude teachers in under-resourced settings, and prior digital experience among participants may have shaped outcomes. Correlational findings, the team stresses, are associative rather than predictive or causal, and the study captured only a single abbreviated DBR cycle rather than the multi-cycle iterations the methodology ideally demands.

Even so, the study sketches a practical blueprint for the future of teacher upskilling. The authors suggest that H5P micro-lessons can serve as digital warm-ups before lessons, AI-generated formative feedback can be embedded into routine assessment, and institutional analytics dashboards can flag teachers who would benefit from targeted coaching. Offline and text-based variants of the modules could extend the model to low-resource environments, and future iterations may incorporate augmented reality, virtual reality and voice-based AI feedback. By aligning measurable competence gains with Sustainable Development Goal 4 on quality education, SmartPD makes the case that short, adaptive, AI-assisted professional development can be both scientifically rigorous and scalable, a combination that has eluded teacher training for decades.

Subject of Research: AI-assisted micro-learning professional development for enhancing teachers' digital competence aligned with the DigCompEdu framework

Article Title: SmartPD: a design-based research model for enhancing teacher digital competence aligned with DigCompEdu

Article References: M., S., & R., A. (2026). SmartPD: a design-based research model for enhancing teacher digital competence aligned with DigCompEdu. Journal of New Approaches in Educational Research, 15(1), Article 5. https://doi.org/10.1007/s44322-026-00052-5

Image Credits: AI Generated

DOI: 10.1007/s44322-026-00052-5

Keywords: teacher professional development, digital competence, DigCompEdu, artificial intelligence, micro-learning, design-based research, H5P, ChatGPT, learning analytics, xAPI, SDG 4, educational technology

The core idea behind AI-TEACH is what the researchers call a co-teacher paradigm. Rather than replacing educator judgment, the system acts as a silent, objective partner that augments it. Strategically positioned surveillance cameras with synchronized audio capture stream classroom data to an edge computing layer, where parallel pipelines analyze behavior and speech in near real time. On the video side, frames are preprocessed with OpenCV and passed through YOLO-NAS, a neural architecture search-optimized object detector that isolates each student in the room. A tracking algorithm called ByteTrack then assigns each detected student a persistent identity across frames, preserving the spatial-temporal coherence needed to study individual behavior over time. MediaPipe Pose extracts skeletal landmarks—nose, eyes, shoulders, wrists, hips—while MediaPipe’s facial model tracks 468 landmark points across the mouth, eyes, and jawline, providing the geometric substrate from which distraction cues are inferred.

From those landmarks, the system derives a structured taxonomy of behavioral distraction. A neck-tilt angle computed between the head axis and torso axis in three-dimensional skeletal space reveals slouching or simulated sleep, especially when combined with prolonged eye closure measured through the eye-aspect ratio and sustained stillness. Wrist-to-hip proximity paired with a downward gaze estimate flags mobile phone use. Rapid arm extension combined with a fast-moving optical flow contour, extracted via the Farnebäck method, identifies thrown objects. Exaggerated facial expressions are caught by measuring the standard deviation of facial landmark displacements across a sliding window of roughly ten frames, and an abrupt displacement of the hip keypoint toward a mapped door location signals a student bolting from class. Each detected incident is logged with a timestamp, an indicator type, and the responsible modality.

The audio pipeline complements the cameras by capturing the verbal disruptions that vision alone cannot see. Incoming sound is first segmented using Silero Voice Activity Detection, a lightweight neural model that separates speech from background noise with low latency. Detected speech segments are then passed through Wav2Vec2, a self-supervised speech model that produces rich acoustic embeddings, which a Support Vector Machine classifies into specific distraction categories: off-task back-talk occurring outside the teacher’s directional audio profile, loud or chaotic sound events such as shouting or whistling identified through amplitude envelope tracking and spectral flatness, rhythmic tapping detected via autocorrelation and short-time energy, and abusive or offensive language flagged by a toxicity classifier applied to transcribed utterances. The design consciously draws on recent advances in multimodal emotion recognition, including per-sample modality equilibration and adversarial fusion techniques, to keep weaker signals from being drowned out by dominant ones when audio and video streams arrive asynchronously.

Fusing these streams is the job of a Bidirectional Long Short-Term Memory network, a recurrent architecture that reads each multimodal feature sequence in both temporal directions, capturing escalating patterns of disruption that a frame-by-frame classifier would miss. A softmax layer over the pooled hidden states assigns each incident a distraction category, and every category carries an empirically calibrated severity weight reflecting its actual disruptive impact on the class—from a single slouching student to thrown objects or hostile verbal outbursts. Over the course of a session, the fog layer multiplies each indicator’s occurrence count by its severity score and sums the results into a single Class Distraction Score, alongside timestamped behavioral metadata. Only this structured metadata, not the raw recordings, is transmitted over encrypted channels to the cloud layer, where an interactive dashboard visualizes the score, breaks down each indicator with its frequency and severity, and plots a temporal heatmap of when distraction clustered during the lesson.

The engineering choices are deliberately pragmatic. The fog layer runs on edge hardware comparable to an NVIDIA Jetson Xavier NX, sustaining frame-level video analytics and concurrent audio processing at 25 to 30 frames per second, while the cloud tier requires only a server-class GPU and at least 10 Mbps of upstream bandwidth. Under these conditions the system achieves an average end-to-end latency of roughly 350 milliseconds from frame capture to logged distraction event, consuming about 15 watts at the edge during continuous operation. Teachers need no technical training beyond an estimated one to two hours of onboarding, because the system is designed for post-session reflection rather than in-the-moment alarms: educators review the dashboard, add contextual notes where needed, and adjust pacing, format, or grouping in subsequent lessons.

The validation combined benchmark testing with a live classroom experiment. The team assembled a curated dataset of more than 5,350 labeled classroom images drawn from three open-source repositories, spanning behaviors such as phone use, object throwing, slouching, yawning, gaze aversion, and head-on-desk posture. The video module achieved 90% overall accuracy and a weighted average recall of 92%, with a macro-average F1 score of 92% across nine evaluation rounds, indicating balanced performance on both frequent and rare distraction categories. The audio module performed comparably, with accuracy between 88% and 94% and a macro-average F1 of 92%, maintaining reliable classification despite overlapping speech and ambient noise. In head-to-head comparisons, AI-TEACH’s overall F1 score of 93.5% substantially outperformed conventional baselines, including a convolutional neural network at 74.14%, a standalone LSTM at 76.04%, HOG plus SVM at 80.52%, OpenPose plus SVM at 83.26%, and even a BiLSTM used alone at 85.95%—evidence that the advantage comes from the fusion of complementary modalities and severity weighting rather than any single component.

The pedagogical test was a controlled experiment with 200 students randomly assigned to experimental and control groups. Control classrooms received traditional instruction, while teachers in the experimental group worked with AI-TEACH’s real-time feedback. After baseline pretests, the experimental group gained an average of 9 points on post-tests compared with 5 points for the control group, and a two-sample t-test confirmed the difference was statistically significant at the 0.05 level. The calculated effect size, Cohen’s d of approximately 0.90, qualifies as large by conventional standards—a striking result for an intervention that changes nothing about the curriculum itself and everything about how quickly teachers can perceive and respond to disengagement. The finding aligns with earlier research showing that objectively measured engagement predicts academic performance better than self-reported distraction.

The authors are candid about the ethical terrain. Because several monitored behaviors—fidgeting, repetitive movement, gaze aversion—can be natural self-regulation strategies for neurodivergent students with autism or ADHD, the system is explicitly designed as a class-level tool rather than an individual diagnostic instrument, and future versions will let educators suppress specific indicators for particular students. Algorithmic bias across skin tones, body types, cultural behavioral norms, and languages remains a recognized limitation, as does the single-institution setting of the 200-student trial and the absence of physiological signals such as heart rate variability. Raw audio and video never leave the edge; only encrypted metadata is stored remotely under role-based access control, with informed consent and periodic bias audits recommended as deployment prerequisites. Framed against the United Nations Sustainable Development Goals, particularly SDG 4 on quality education, AI-TEACH represents a broader shift toward classrooms where attention itself becomes measurable, and where teachers—armed with evidence instead of intuition—can reach drifting students before the drifting becomes permanent.

Subject of Research: Multimodal AI-based real-time assessment of student distraction and engagement in classrooms

Article Title: Artificial Intelligence Based Framework for Student Engagement Assessment in Classroom Environments

Article References: Saini, M., Sharma, H., & Sengupta, E. (2026). Artificial Intelligence Based Framework for Student Engagement Assessment in Classroom Environments. Cognitive Computation, 18(1), Article 108. https://doi.org/10.1007/s12559-026-10629-z

Image Credits: AI Generated

DOI: 10.1007/s12559-026-10629-z

Keywords: artificial intelligence, classroom engagement, distraction detection, computer vision, speech recognition, BiLSTM, fog computing, education technology, multimodal fusion, smart education, SDG 4, student behavior