Search Results (6,592)

Large classes in Chinese College English programmes make frequent analytic assessment of student writing difficult. Large language models (LLMs) may support more frequent formative assessment, but their scores may vary across queries and be systematically harsher or more lenient than local teacher ratings. Using a corpus-based, five-fold cross-validated comparative rater-evaluation design, this study examined whether statistical calibration could make LLM-assisted scores more interpretable for College English writing assessment and where their use should remain limited. Data comprised 414 timed argumentative essays written by Chinese non-English majors at one applied undergraduate institution. Two trained College English teachers independently rated the essays on a seven-dimension analytic rubric informed by China’s Standards of English Language Ability, providing the local reference standard. Three LLMs rated each essay–dimension pair on five occasions. Under five-fold cross-validation, uncalibrated scores were compared with location–scale correction, isotonic calibration, and equipercentile linking, using quadratic weighted kappa, Spearman correlation, mean absolute error, signed bias, and half-point tolerance accuracy. Agreement between models did not imply agreement with teachers: two models showed inter-model kappa values of 0.70–0.78 but an average kappa of only 0.15 with teacher ratings while rating the essays about one band more severely. Calibration removed most of this severity difference and raised pooled kappa to 0.61–0.70 depending on the method (0.63–0.64 under equipercentile linking), compared with a teacher–teacher agreement benchmark of 0.747. The three methods differed little, and the improvement mainly reflected closer alignment of score distributions rather than better judgement of writing quality. Agreement was higher for vocabulary, syntax, and grammar but remained low for cohesion and conventions. The findings suggest that LLM-assisted scoring may support low-stakes formative feedback when calibrated to local teacher standards and used under teacher supervision, while teachers retain responsibility for judging content, coherence, argumentation, and communicative quality. Full article

Deliberate practice plays a central role in high-quality music learning, yet the motivational and technology-related factors associated with sustained deliberate music practice remain insufficiently understood. Drawing on social cognitive theory, achievement goal theory, and the literature on artificial intelligence (AI) literacy, this study examined the relationships among music learning self-efficacy, achievement goal orientation, AI literacy, and deliberate music practice. A total of 458 music students from a teacher-training university in Liaoning Province, China, completed measures of the four constructs. Music learning self-efficacy was positively associated with deliberate music practice. Achievement goal orientation also showed a significant indirect association between music learning self-efficacy and deliberate music practice, such that students with higher self-efficacy reported stronger achievement goal orientations, which in turn were associated with greater engagement in deliberate music practice. AI literacy further moderated the association between achievement goal orientation and deliberate music practice, with this positive relationship being stronger among students reporting higher levels of AI literacy. These findings suggest that deliberate music practice is associated with both learners’ motivational beliefs and their self-reported AI literacy. Full article

The diffusion of generative artificial intelligence has moved educational debate beyond access to new tools, placing teacher work, pedagogical mediation, assessment, and digital competence under renewed pressure. This article examines teachers’ perceptions, declared uses, and forms of engagement with generative AI in Gymnasien (academic-track secondary schools) and Gesamtschulen (comprehensive secondary schools) in North Rhine-Westphalia, Germany. The study is based on an anonymous and voluntary online questionnaire circulated through school leaderships. The operational database comprised 204 records, while the main analysis was based on 156 completed questionnaires. Closed-item responses were examined through descriptive statistics and interpreted through DigCompEdu and its AI-related supplement. The findings indicate an asymmetrical incorporation of generative AI across teachers’ work. Its presence is stronger in preparation, organisation, professional reflection, and risk awareness than in direct classroom mediation, process-oriented assessment, and institutionally supported student use. The strongest contrasts concern perceived student overreliance on AI-generated answers, the need to adapt assessment practices, limited school-level guidance, and weak use of AI-supported process monitoring. Rather than interpreting this imbalance as a simple delay in adoption, the article argues that teacher competence in relation to generative AI is produced through the interaction between professional judgement, pedagogical purposes, and institutional conditions. It therefore cannot be reduced to an individual technical attribute. Full article

Education for sustainable development (ESD) has gained increasing attention over the last few decades, as human innovation and prevailing consumer practices have driven significant environmental degradation, resource depletion, and social inequalities. ESD must extend beyond the transmission of knowledge; it must shape mindsets, behaviors, and systems. Accordingly, fostering sustainability competencies among primary and secondary school students has become an educational imperative. The aim of the study was to identify opportunities for developing eight key competencies for advancing sustainability transformations within obligatory and elective chemistry curricula in Slovenian lower and upper secondary schools. A qualitative content analysis of chemistry curriculum documents was conducted, focusing on general learning objectives, operational learning objectives, and associated learning outcomes. The results indicate that general learning objectives of the obligatory and elective chemistry curriculum for lower and upper secondary educational levels enable opportunities for the development of all key competencies for advancing sustainability transformation, whereas the extent to which these competencies can be developed through operational learning objectives and learning outcomes varies across chemistry curricula and educational levels. However, despite the existing possibilities in curricula, teachers play a crucial role in their implementation in classrooms. Therefore, it is essential to support teachers in recognising the importance of developing these competencies and meaningfully selecting didactic approaches that foster the development of all key competencies for advancing sustainability transformation within the subjects they teach during pre-service and in-service education. Full article

Correcting student misconduct is a central issue in educational psychology and classroom management; teachers often use punishment to correct students’ inappropriate behavior. Numerous empirical studies have shown that punishment influences individuals’ willingness to change their behavior, but the underlying mechanisms remain to be elucidated. This study employed a scenario simulation method to survey 240 middle school students, aiming to explore the relationship between types of punishment and students‘ self-reported willingness to change their behavior in response to hypothetical scenarios. Furthermore, the study examined the chained mediating effects of perceived punishment motivation and trust, as well as the moderating effect of group relationship. The results showed that perceived punishment motivation and trust exert a chained mediating effect between punishment type and self-reported willingness to change behavior. Punishment first is associated with enhanced perception of the prosocial motivation behind the punishment, which in turn is associated with higher trust in the punishing teacher, and ultimately with greater self-reported willingness of misbehaving students to change their behavior; that is, perception of punishment motivation and trust sequentially transmit the effects of punishment. Furthermore, this transmission process is moderated by group relationship: when the punisher is a teacher (in-group), the influence of punishment type on students’ self-reported willingness through perceived punishment motivation and trust is stronger, whereas this effect is not significant when the punisher is an AI (out-group). These findings reveal the psychological pathway through which punishment influences self-reported willingness to change behavior in a hypothetical scenario via sequential processing of cognitive attribution (perceived punishment motivation) and attitudinal evaluation (trust), and highlight the moderating role of group relationship in this process. Full article

External resources, digital competence, and psychological characteristics have each been linked to Artificial Intelligence (AI) literacy, but almost always in isolation, so the unique contribution of any one domain remains unknown. Grounded in social cognitive theory, this study models all three domains jointly. Data were collected from 303 undergraduates at a liberal arts university in Hong Kong and analysed using hierarchical multiple regression. Perceived resources and perceived teacher support explained 29% of the variance in AI literacy; prior digital competence added a further 3%; and personal innovativeness, technology anxiety, and growth mindset in technology added a further 6%, with all three being significant in the final model (total R² = 0.39). The external predictors remained significant throughout but attenuated substantially, indicating that institutional provision is necessary but not sufficient. Technology anxiety, which correlated negatively with AI literacy at the zero-order level, emerged as a positive predictor once other determinants were controlled. Analyses of the five AI literacy subdimensions localised this effect to critical evaluation and ethical competence and found no association with the three performance-oriented dimensions, suggesting that anxiety operates through heightened vigilance rather than enhanced operational skill. These findings, which no single-domain design could have produced, indicate that fostering AI literacy requires attention to students’ psychological readiness and foundational digital skills alongside the provision of infrastructure. Full article

Region-level weakly supervised hyperspectral target detection (HTD) using multiple-instance learning (MIL) reduces annotation costs but encounters challenges such as bag label ambiguity, boundary over-smoothing, and test-time computational latency. To address these issues, we propose Hyper-VMIL, a spatial–spectral topology-regularized variational hypergraph network. Hyper-VMIL formulates latent target localization as variational inference over dual-path hypergraphs: a boundary-aware spatial hypergraph modeling geometric patch continuity and a dynamic spectral-manifold hypergraph capturing non-local material similarity. Node-adaptive gating dynamically balances spatial and spectral evidence to mitigate over-smoothing near target boundaries. Furthermore, a confidence-aware continuous posterior refinement (CTPR) mechanism reduces the confirmation bias associated with conventional hard pseudo-label binarization. Finally, a teacher–student distillation strategy transfers contextual topology into a lightweight single-spectrum student detector. Benchmark experiments on simulated ASTER and airborne MUUFL Gulfport and Avon datasets show that Hyper-VMIL achieves competitive performance against 15 baseline methods. Notably, Hyper-VMIL supports dual inference modes: Context Mode provides improved detection accuracy (+4.6% average NAUC over VMIL-ECM on MUUFL), while Pixel Mode enables single-spectrum inference ($1.25\mathsf{\mu }\mathrm{s}$ single-instance latency and an amortized streaming throughput of $0.015\mathsf{\mu }\mathrm{s}$ per pixel) suitable for onboard real-time deployment. Full article

This exploratory descriptive-correlational study analyses the perceptions of 257 adult students (doctoral, master’s, and teachers) at DPPD, USV Iași, during the 2025–2026 academic year regarding hybrid education, teaching quality, and environmental sustainability. Using a structured Likert-scale questionnaire, the analysis indicates good-to-excellent internal consistency, with Cronbach’s alpha values ranging between 0.886 and 0.901, and a high overall global average score of 4.64. The findings reveal strong support for the hybrid model. Perceived teaching quality received the highest subscale rating (M = 4.80), closely followed by the perceived ecological impact (M = 4.65). The analysis indicates strong Pearson correlations, specifically between the hybrid learning experience and perceived teaching quality (r = 0.818), as well as between the perceived ecological impact and pro-sustainability attitudes (r = 0.809). Regarding academic mobility, the estimate indicates 81,283 km of avoided commuting travel and approximately 12,295 kg of avoided commuting-related CO₂ emissions, based on self-reported distance, means of transport, and number of physical attendances replaced by online activities. These findings suggest that hybrid learning may represent a relevant managerial option for university sustainability policies. The model appears well suited to postgraduate programmes addressed to employed adults, although the ecological benefits should be read as partial and do not displace perceived teaching quality as the central factor. Full article

Plant disease symptoms combine local texture changes with patterns distributed across a leaf, while practical recognition models must remain compact. We introduce DC-FEN, a MobileNetV3-based design that models spatial-token relations and channel interactions in parallel and injects them through gated residual fusion. We also examine output-distribution, direct-feature, and token-relation transfer under same-backbone and heterogeneous teachers. PlantVillage and Plant Pathology 2021 (FGVC8) are evaluated with duplicate-audited, group-aware 70/15/15 splits, an explicit unresolved-leaf sensitivity check, validation-only selection, five training seeds, class-sensitive metrics, and paired seed-wise descriptive summaries. On PlantVillage, the no-additional-attention student, DC-FEN teacher, and DC-FEN joint student obtain macro F1 scores of $96.46\pm 0.91\%$, $96.90\pm 0.40\%$, and $96.55\pm 0.25\%$. On FGVC8, the corresponding scores are $87.29\pm 0.63\%$, $87.14\pm 0.52\%$, and $87.20\pm 0.26\%$. At the prespecified FGVC8 threshold of 0.5, DCAB changed sample-wise F1 by $-0.02\pm 0.55$ percentage points relative to the unmodified backbone; validation-selected global and label-specific thresholds changed this contrast to $+0.28\pm 0.55$ and $+0.55\pm 0.29$ points, while threshold-free macro mAP remained essentially unchanged. A duplicate-audited PlantDoc pressure test reduced frozen-checkpoint accuracy to $30.34\pm 1.10\%$ and $29.57\pm 1.10\%$, showing that external generalization remains unestablished. A ResNet50 teacher gives logit-only students $97.42\pm 0.51\%$ macro F1 on PlantVillage and $89.82\pm 0.43\%$ sample-wise F1 on FGVC8. After separately weighting the direct and relation terms, the corresponding joint students obtain $97.37\pm 0.56\%$ and $89.94\pm 0.27\%$, recovering the degradation seen with unit internal weights while remaining close to logit-only transfer. Thus, the study evaluates the benefits and limits of explicit spatial–channel interaction and shows that adding intermediate transfer constraints does not guarantee a stronger student. Full article

With the development of smart classrooms and digital learning platforms, multimodal learning behavior data provide a new sensing basis for understanding students’ cognitive states and psychological risk warnings. However, existing educational data mining methods mainly focus on performance prediction, dropout warning, or surface-level emotion recognition, while continuous and interpretable modeling of deeper cognitive biases and related psychological risks remains insufficient. To address this issue, we propose MLBS-Net, a multimodal learning behavior sensing network for teaching feedback that jointly models students’ textual expressions, behavioral sequences, classroom interactions, and psychological auxiliary signals. MLBS-Net integrates theory-guided textual cognitive bias encoding, temporal behavioral state modeling, and reliability-aware cross-modal fusion to capture psychologically interpretable cognitive patterns, characterize dynamic learning-state changes, and adaptively integrate multimodal information according to data quality and task contribution while providing interpretable feedback for teachers. Experimental results show that MLBS-Net achieves a Macro-F1 of 0.855 for cognitive bias recognition and an AUC of 0.891 for psychological risk warning, outperforming traditional machine learning, unimodal deep learning, and standard multimodal methods. Ablation results further support the effectiveness of theory-guided semantic encoding, temporal behavioral modeling, reliability estimation, and multitask learning. These findings demonstrate that MLBS-Net can jointly characterize cognitive biases and potential psychological risks from multisource learning behaviors, providing a feasible approach for learning-state sensing, risk warning, and interpretable teaching support in smart education. Full article

This study examines the role of generative artificial intelligence in higher education, focusing specifically on creative degree programs and students’ perceptions of its academic and creative value. Employing a mixed-methods design, data were collected from 555 university students enrolled in communication- and design-related degrees in Spain. By combining an online survey with focus groups, the research analyzed the frequency, purposes, and meanings of AI use in academic tasks. The results show that students primarily use generative AI to clarify concepts, develop ideas, review literature, and support academic production. Although they acknowledge its utility as a learning tool, participants also raised concerns regarding overreliance, reduced creative effort, unreliable outputs, and potential threats to originality and authorship. Consequently, the study concludes that while generative AI is already influencing learning practices in higher education, its educational potential relies heavily on clear pedagogical guidance, ethical implementation, and active teacher mediation. Based on these findings, the article proposes a ten-step teaching framework for AI-supported creative interactive content design, aimed at fostering pedagogical innovation while preserving critical thinking, creativity, and student authorship. Full article

Background/Objectives: This study used structural equation modeling (SEM) to examine the sequential mediating roles of school belonging and prosocial behavior in the effects of parent–adolescent and teacher–student relationships on adolescents’ academic achievement. Methods: Using a cross-sectional design, the study included 523 adolescents (61.6% female, 38.4% male) aged 13 to 19. Data were collected with the Mother–Adolescent Relationship Quality Scale, the Teacher–Student Relationship Quality Index, the School Belongingness Scale, and the Student Prosocial Behavior Scale, together with subjective (perceived achievement) and objective (school grades) indicators of academic achievement. Results: Both parent–adolescent (β = 0.25) and teacher–student (β = 0.44) relationships significantly predicted school belonging; school belonging significantly predicted prosocial behavior (β = 0.22); and prosocial behavior, in turn, significantly predicted academic achievement (β = 0.22). The direct effect of parent–adolescent relationships on academic achievement was significant (β = 0.26), whereas neither the direct effect of teacher–student relationships (β = 0.13) nor the direct effect of school belonging on academic achievement was significant. The indirect effects of both relational contexts on prosocial behavior via school belonging were significant, and school belonging predicted academic achievement indirectly through prosocial behavior; however, neither the total indirect effect of parent–adolescent relationships on academic achievement, nor its specific serial indirect effect through school belonging and prosocial behavior, reached statistical significance. The corresponding specific serial indirect effect for teacher–student relationships was statistically significant, although small in magnitude and not accompanied by a significant total indirect effect. Conclusions: These findings indicate that school belonging is more closely tied to socio-emotional processes than to academic achievement directly, and that the associations between relational contexts and academic achievement followed a differentiated rather than a uniform pattern: parent–adolescent relationships were linked to achievement primarily through a direct association, and their hypothesized distal indirect pathway to achievement via school belonging and prosocial behavior was not statistically supported. Teacher–student relationships, by contrast, showed no significant direct association with achievement, but the corresponding distal indirect pathway was statistically supported, although the effect was small and imprecisely estimated, underscoring the need for cautious interpretation of these relational mechanisms. Full article

Team sports are an important component of physical education (PE), and teachers’ instructional decisions shape students’ learning experiences and outcomes. This study examined associations between teacher age and didactic preferences in team sports pedagogy among 1856 Slovak PE teachers. A cross-sectional survey design was used, and participants were categorized into three age groups (≤30 years, 31–50 years, and ≥51 years). Associations between age and didactic preferences were analyzed using chi-square tests, adjusted standardized residuals, and ordinal and multinomial logistic regression models. Significant age-related differences were identified for didactic approach (χ²₍₄₎ = 14.67, p = 0.005, V = 0.063), teaching style (χ²(4) = 15.25, p = 0.004, V = 0.064) and methodological-organizational form (χ²(8) = 21.71, p = 0.006, V = 0.076), although effect sizes were small. Teachers aged ≤30 years had higher odds of preferring a technical rather than combined approach than teachers aged 31–50 years (OR = 1.958, 95% CI = 1.344–2.852, p < 0.001) and teachers aged ≥51 years (OR = 1.563, 95% CI = 1.019–2.399, p = 0.041). Teachers aged 31–50 years showed stronger preferences for practical rather than command teaching styles than both younger (OR = 1.455, p = 0.018) and older teachers (OR = 1.386, p = 0.023). Pearson’s chi-square test did not identify significant associations with teaching-unit duration (p = 0.052); however, ordinal logistic regression indicated significant overall age-group effects (p = 0.022). No significant age-related association was observed for preferred game activity (p > 0.05). The findings indicate that teacher age is associated with selected didactic preferences in team sports instruction. Although the observed associations were small, they may reflect variation in professional socialization and teaching experience and may have implications for PE teacher education and professional development. Full article

Project-based learning (PBL) holds significant promise as a student-centered instructional approach in secondary mathematics, yet teachers’ capacity to leverage it for diverse learners is deeply shaped by their affective experiences during implementation. This narrative inquiry examines the affective journeys of three high school mathematics teachers as they adopted PBL in classrooms characterized by varied student readiness levels and engagement profiles. We document teachers’ affective journeys as they experience initial PBL awareness, participate in active experimentation, and work toward forming emerging commitments and facilitator identities. Analysis of focus groups and interviews reveals three affective tensions: comfort versus transformative reflection, epistemic ideals versus ontological school realities, and career risk versus identity change. The findings demonstrate how teachers can navigate iterative methods of trial, how teacher educators can better support future teachers’ use of student-centered pedagogies, and how administrators can be aware of the impact that teacher affect can have on performance and pedagogy. Implications address how this broader education community can support differentiated, inquiry-based mathematics instruction. This support can be advanced through scope-controlled projects, micro-evidence portfolios, and incremental adoption pathways. These approaches honor teachers’ emotional labor while supporting their sustained use of student-centered learning strategies intended to broaden opportunities for participation and learning in mathematics classrooms. Full article

The Cambodian Ministry of Education Youth, and Sport (MoEYS) has, over time, made noticeable progress toward Sustainable Development Goal 4 (SDG-4) to ensure inclusive and equitable quality education and promote lifelong learning opportunities for all. However, the pace and extent of that progress remain uneven, particularly in the area of teacher training, which is central to achieving inclusive and quality education. Utilising a document analysis framework, this study draws upon national policy reform documents including the Teacher Policy Action Plan to examine the extent to which the MoEYS’s policy commitments to professionalise the teaching workforce are meeting SDG-4.c goals. Changes to pre-service programmes including the 12 + 4 programme, investments in teacher training centres, and continuous professional development (CPD) frameworks have strengthened the system. Yet, significant challenges persist. Limited access to effective CPD, high student/teacher ratios, teacher shortages, especially in rural locations, and resource constraints continue to undermine the consistency and quality of instructional practice. A contention of this paper is that for Cambodia to fulfil its SDG-4.c target relating to teacher preparedness, progress is contingent upon sustained investment, targeted support for rural and remote regions, and stronger implementation capacity to ensure all learners benefit from well-trained and effectively supported educators. Full article