Search Results (392)

This study examines the increasing concern regarding teacher job satisfaction, which has a direct impact on retention, instructional quality, and student outcomes. Traditionally, teacher satisfaction has been evaluated through questionnaires, which present limitations in terms of data efficiency and analyses. In this study, machine learning techniques were applied to data from the PISA 2022 teacher questionnaire in Morocco (N = 2998 lower-secondary teachers). Two multiclass classification targets were defined: overall job satisfaction (SATJOB_class) and satisfaction with the teaching profession (SATTEACH_class), each categorised into three balanced classes: low (<−0.5), medium (−0.5 to 0.5), and high (>0.5) classes. The methodology comprised four key stages. Initially, comprehensive pre-processing was conducted to address missing values, retaining features with fewer than 300 missing entries and applying mode imputation. Subsequently, nine classifiers, including logistic regression, K-nearest neighbours, multinomial naïve Bayes, support vector machine, decision tree, random forest, XGBoost, AdaBoost, and a feed-forward Artificial Neural Network, were evaluated using identical train/test splits and hyperparameter tuning. Third, the model performance was assessed using accuracy, precision, recall, and F1-score. Finally, the feature importance was derived from tree-based and permutation methods. The results indicated that XGBoost outperformed the other models for SATJOB_class with an accuracy (0.61), precision (0.62), recall (0.61), and F1-score (0.61), followed by Random Forest (accuracy = 0.59), Logistic Regression (accuracy = 0.59), and AdaBoost (accuracy = 0.59). For SATTEACH_class, Random Forest led with accuracy (0.59), followed closely by XGBoost (0.58), ANN (0.57), and AdaBoost (0.56). Key predictors of teacher job satisfaction included workload-related variables and school-environment factors, which consistently emerged as the most important features across the best-performing models. The methodology and open-source pipeline provide a reproducible framework for evidence-based interventions to improve teacher retention and instructional quality, offering valuable insights for policymakers and educational administrators. Full article

Knowledge distillation (KD) transfers knowledge from large language models (LLMs) to smaller or similarly sized models in order to obtain efficient yet capable systems. However, performing distillation over all tokens is computationally expensive and may weaken the transfer signal. To address this limitation, Knowledge-Filtered Distillation (KFD) is introduced as a selective distillation approach in which tokens are filtered according to the divergence $KL({\mathcal{M}}_2\mid {\mathcal{M}}_0)$ between a teacher model (${\mathcal{M}}_2$) and a base model (${\mathcal{M}}_0$), while the student model (${\mathcal{M}}_1$) is also derived from the same base model. Only tokens whose divergence exceeds a predefined threshold are distilled. For the selected tokens, the teacher distribution is normalized over the Top-5 predictions, whereas tokens outside this case receive a label-ranking bonus. The proposed conditional Top-5/bonus target design is shown theoretically to yield a lower label-focused target error than using only Top-5 normalization or only the bonus across all tokens. In addition, the KL and cross-entropy (CE) losses are balanced through a dynamically computed batch-level coefficient $\alpha$. Experiments on multiple Turkish text datasets show that KFD consistently outperforms CE-only training, achieving higher accuracy with less data and shorter training time. KFD also outperforms entropy-based token selection methods and highlights the role of student initialization in effective knowledge transfer, thereby providing an efficient and scalable distillation framework for teacher–student models of equal size. Full article

A system of initial teacher education must incorporate instruments capable of capturing the complexity of professional learning. In this context, national diagnostic assessments have become central mechanisms for monitoring outcomes in initial teacher education. This study examines student performance in Initial Teacher Education in Physical Education (ITEPE) programs using the themes assessed by the National Diagnostic Assessment (NDA) 2024, comparing achievement levels and analyzing differences according to the type of higher education institution in Chile. A quantitative, cross-sectional, and comparative design was employed, using official data from 1102 students enrolled in all Chilean universities offering the program. Descriptive and nonparametric inferential analyses were conducted to examine differences by sex and type of institution. The results show relatively homogeneous performance across standards, with higher percentages of achievement in dimensions related to didactic organization and assessment, and lower results in the standard associated with understanding student characteristics. Although statistically significant differences were identified according to administrative dependency, the effect sizes were small. Consequently, the NDA is positioned as a formative input to guide contextualized curricular improvements rather than as a mechanism for institutional ranking. Full article

The urgency of global sustainability challenges increased policy attention to Education for Sustainable Development (ESD), particularly in relation to Sustainable Development Goal 4.7, which calls for the systematic integration of sustainability competences across education systems. This article examines how organisational and structural approaches to continuous professional development (CPD) can support the institutionalisation of ESD beyond individual teacher training. The article adopts a case-based analytical approach drawing on programme documentation and evaluation data from two initiatives coordinated by the teacher training department of the District Government of Arnsberg in North Rhine-Westphalia, Germany: the Erasmus+ consortium EFFORT-A, which links international mobility with school development processes, and the regional programme WIRkstatt Zukunft, which implements the Whole School Approach through modular training and school-based consultancy. The analysis indicates that multi-level governance, structured networking, leadership engagement, and formal contracting mechanisms are associated with the integration of ESD within school cultures, curricula, and organisational routines. Challenges are identified regarding resource allocation, policy coherence, and the long-term sustainability of project-based formats. The article concludes that sustained ESD implementation requires CPD systems that combine international perspectives with regionally anchored support structures and align individual professional learning with institutional development strategies, offering recommendations for policymakers and educational leaders. Full article

We introduce Self-Competitive Distillation (SCD), a parameter-neutral training strategy aimed at influencing optimization dynamics without increasing model size or relying on external teachers. Two identical instances of the same architecture, initialized with different random seeds, are trained jointly and dynamically exchange asymmetric teacher–student roles based on instantaneous performance, enabling knowledge transfer between diverging optimization trajectories. Under fixed parameter and training budgets, SCD is observed to improve the realized effective capacity of lightweight architectures, yielding a higher test accuracy at matched epochs. Across multiple lightweight vision models and datasets, SCD demonstrates gains in both in-domain performance and cross-domain generalization, as measured by xScore. These results suggest that, within the evaluated experimental conditions, SCD can help mobile models make more effective use of training dynamics, while the underlying architecture remains the primary determinant of effective capacity in resource-constrained settings. Full article

As artificial intelligence (AI) drives significant challenges in education, understanding and addressing the training needs of in-service teachers has become a critical issue for ensuring a responsible and long-term technological transition. Framed within Sustainable Development Goal 4 (SDG 4) and the principles of Education for Sustainable Development (ESD), teacher preparation in AI is increasingly recognized as a key mechanism for promoting ethical, equitable, and inclusive educational transformation. This study explores the influence of several key variables on intention and learning behaviors in relation to AI among a sample of 704 Spanish in-service teachers (71% women) from all compulsory educational levels. Using a validated questionnaire, this study assessed teachers’ anxiety towards AI, basic AI knowledge, personal relevance of AI, AI for social good, perceived self-efficacy, social pressure, and perceived usefulness of AI. Structural equation modeling (SEM) was employed to analyze the direct and indirect relationships among these variables. The results indicate that the perceived usefulness of AI and self-efficacy directly and positively influence the behavioral intention to learn about AI. Furthermore, social pressure and basic AI knowledge indirectly influence this intention. In turn, both behavioral intention and social pressure significantly predicted AI learning behaviors. The model demonstrates strong explanatory power, accounting for 91% of the variance in the behavioral intention to learn about AI. These findings provide evidence to inform the design of teacher training initiatives and policies that promote responsible, ethical, and inclusive integration of AI in educational settings, contributing to sustainable development through education. Full article

Cross-sectional surveys of psychological distress using the Kessler-6 tool (K6+) were conducted among training cohorts per year of New Jersey (NJ) secondary school teachers between January 2022 and December 2024. Data downloaded for 12–18 annual virtual synchronous live session training date ranges related to specified teacher cohorts, consisting of 30 calendar days prior to its date to relate to K6+ questions (575 unique participants across 42 total live sessions). Utilizing data from federal/state air quality monitoring stations (AQMS), we constructed a database of estimated exposures to ambient/outdoor air quality. Cohorts were broken down by school district (SD) and paired with AQMS based on approximate geographic proximity for each SD’s school’s physical address utilizing NJ-GeoWeb. Once addresses were reported and associated with two AQMS, associated reviewed daily criteria pollutant data (2021–2024) were retrieved for particulate matter (PM, PM₁₀ and PM_2.5) and ozone. Data were averaged for relevant stations. Analyses suggested prior 30-day PM_2.5 showed a significant negative correlation with K6+ scores, −0.32 with PM_2.5 concentration (p = 0.04) and −0.48 with PM_2.5 AQI (p = 0.002); however, wind speed had a positive association, 0.33, with K6+ scores (p = 0.03). These results suggested how specific events and meteorological conditions affected ambient air quality for only some of the prior 30 days yet still potentially influenced K6+ scores for some cohorts, e.g., large wildfires then prevailing winds. More research with improved exposure assessment is warranted. This initial environmental epidemiology study with ecological design can inform future collaborative research and practice work on mental health and the effects of environmental factors. Full article

This study explores the effectiveness of Baskin—an inclusive sport discipline—as a pedagogical tool to promote school inclusion, social cohesion and motivation in secondary school students. The intervention, conducted on a sample of 600 students in four Italian institutions, adopted an experimental design with pre–post measurements and a mixed methods approach. The quantitative results, obtained by means of validated psychometric instruments (PIQ, Classroom Cohesion Scale, AMS), showed significant improvements in all variables investigated in the experimental group compared with the control (p < 0.001). The qualitative analysis, based on interviews, focus groups and reflexive diaries, highlighted five thematic areas: revaluation of diversity, improvement of the classroom climate, development of self-efficacy, restructuring of interpersonal relationships and request for project continuity. Baskin emerged as a comprehensive educational practice, capable of integrating corporeity, citizenship and critical thinking. The systematic inclusion of inclusive sport in the curriculum and initial teacher training is suggested, as well as the promotion of school networks and longitudinal studies. Baskin emerges as a pedagogy of participation that contributes to transforming school culture, making inclusion concrete. Full article

In recent years, deep learning models have exhibited exceptional performance across several tasks. However, the substantial computational and storage demands impede implementation on edge devices with constrained resources. Online Knowledge Distillation (OKD) has arisen as an effective model compression strategy that removes the reliance on pre-trained teachers characteristic of conventional distillation approaches. Nonetheless, OKD persists in facing challenges, including substantial performance variances within student networks, insufficient learning capacity for difficult data, and network homogeneity. To address those issues, this paper proposes an Elite-Student-Enhanced Adaptive Online Knowledge Distillation (ESAKD) algorithm. ESAKD introduces a patience factor-based adaptive temperature scheduling mechanism to dynamically balance knowledge clarity and richness during knowledge transfer. This mechanism utilizes the performance benefits of elite-students, particularly during initial training phases, to offer superior supervision that successfully transcends the learning capacity limitations of current OKD approaches. This method promotes swift convergence and substantially enhances the ultimate precision of the standard-student models. Furthermore, a confidence-weighted ensemble student model is designed to improve collective decision-making. Experimental assessments indicate that ESAKD provides substantial performance improvements over supervised learning without distillation and other leading online distillation techniques. On the CIFAR-100 dataset, ESAKD improves the test accuracy of various models by 1.49–6% over the undistilled baselines, and by 0.27–2.18% compared to advanced online distillation algorithms. Moreover, it exhibits enhanced performance on the Tiny-ImageNet-200 dataset as well. Full article

Folklore, as an expression of intangible cultural heritage, plays a key role in the transmission of collective memory, values, and cultural identity and has been widely recognized for its educational potential in early childhood. This study analyzes the perceptions of Early Childhood Education teachers in the province of León (Spain) regarding the pedagogical value of folklore and its use in classroom practice. A quantitative, cross-sectional, non-experimental design was employed using a validated questionnaire administered to a voluntary sample of 100 teachers from schools offering the second cycle of Early Childhood Education. The instrument assessed three dimensions: knowledge of folklore, perceived relevance of folk literature, and application in educational practice. The results indicate that teachers hold very positive attitudes toward folklore as a cultural and educational resource, particularly for children’s social, emotional, and cultural development. However, a clear gap emerges between this strong theoretical appreciation and its occasional classroom use. Significant differences were identified according to school environment, type of school, and teaching experience, indicating that contextual and professional factors influence folklore integration. Insufficient training and low self-confidence were identified as the main barriers. This study highlights the need to strengthen initial and in-service teacher education in cultural heritage and in the didactics of literature and music to promote a more meaningful and sustainable use of folklore in early childhood education. Full article

Educational inclusion remains a global challenge, with teachers’ attitudes toward disability being crucial for the implementation of inclusive classroom practices. This study compares the attitudes toward disability of 252 teachers from Spain and Ecuador, two countries whose educational systems promote inclusion but differ in the maturity of their legislative and institutional frameworks. The aim was to examine cross-national differences in attitudes and analyze how personal and professional variables relate to these attitudes. A descriptive, observational, cross-sectional design was used. Data were collected through an online administration of the Attitudes Toward Persons with Disabilities Scale, Form G via non-probabilistic. The instrument assesses five attitudinal dimensions: capacity assessment, rights recognition, personal involvement, generic rating, and role assumption. Results showed that Spanish teachers reported significantly more positive attitudes than Ecuadorian teachers in all dimensions except role assumption. Among sociodemographic and professional variables, only Early Childhood Education training and prior experience working with individuals with disabilities consistently correlated with more favorable attitudes, while age and teaching experience demonstrated weak and inconsistent associations across countries. These findings underscore the influence of cultural, educational, and institutional contexts on teachers’ attitudes toward disability and highlight the need to reinforce both initial and ongoing preparation in inclusive education. Strengthening structured practicum experiences, socio-emotional competencies, and inclusion-focused coursework may contribute to more positive attitudes, while adapting training policies and inclusive practices to each country’s cultural characteristics is essential for effective and sustainable implementation. Full article

Artificial intelligence (AI) has reshaped contemporary approaches to teaching, assessment, and feedback. Most AI systems provide reactive feedback, offering instant answers that reduce learners’ cognitive engagement and sense of agency. In contrast, Mili was developed as a proactive pedagogical intelligence that asks guiding questions and encourages learners to construct their own responses. Through this design, feedback becomes a process of learning rather than an evaluative mechanism. Mili is a Hebrew-language educational chatbot grounded in principles of dialogic feedback, pedagogical mediation, and literacy resilience. Its goal is to create a metacognitive literacy dialogue in which questions replace answers and learning becomes an act of reflection and self-inquiry. The development followed a Design-Based Research approach involving iterative cycles of design, training, and testing. At each stage, pedagogical prompts were crafted to simulate authentic teacher–learner dialogue, including clarifying questions, pedagogical delay, and emotional reinforcement. This process enabled an exploration of how AI can mediate feedback that stimulates deeper cognitive engagement. The resulting model demonstrates proactive dialogic feedback in which AI does not simply respond but initiates reflective dialogue. Simulated interactions with Mili reveal how such feedback supports the three dimensions of literacy resilience: linguistic-cognitive, metacognitive, and emotional. Mili represents a conceptual shift in AI-based feedback, moving from response to process, from outcome to mediation, and from reactive AI to learning-generative AI. The study makes a theoretical contribution by articulating a model of pedagogically mediated AI and a practical contribution by developing a feedback tool that fosters inquiry, reflection, and literacy resilience in learners and teachers. Full article

The teaching profession increasingly demands complex competencies, including collaboration, professional networking, and adaptability, beyond subject-specific knowledge, due to rapid educational, technological, and labor market changes. This study addresses the limited national data on the professional relational capital of teacher education students and examines how relational networks affect professional identity, commitment, and retention. A pilot questionnaire was developed from focus group interviews conducted in spring 2024 at the University of Nyíregyháza and analyzed with ATLAS.ti 7. The instrument includes four dimensions: parental influence, initiative during high school, initiative during university, and future employment plans, with indicators such as place of residence, cooperation patterns, network durability, domestic and international collaborations, and professional aspirations. Results indicate that students’ relational networks are central to early professional socialization and engagement in collaborative teaching communities. Mapping these networks offers diagnostic and developmental insights, supporting targeted mentoring, inter-institutional cooperation, and international mobility. Findings suggest that deliberately developing relational capital during teacher training enhances professional preparedness, satisfaction, and retention. Overall, the study highlights the value of integrating professional networking and collaborative competencies into teacher education to promote sustainable career paths, align training with labor market expectations, and strengthen the quality, resilience, and long-term sustainability of the teaching workforce. Full article

The integration of digital technologies into early childhood education extends beyond mere technical necessity; it constitutes a fundamental pillar of social sustainability within the teaching profession. Yet, a persistent paradox remains in teacher education: the “Attitude–Competence Gap,” where pre-service teachers’ enthusiasm for technology fails to translate into practical proficiency. This study interrogates this disconnect within a STEAM framework, specifically examining whether digital competence is driven by general technological attitudes or domain-specific pedagogical beliefs. Utilizing an explanatory sequential mixed-methods design, we analyzed data from 200 Child Development students, followed by in-depth semi-structured interviews with 15 participants who exhibited high attitudes but low initial competence. Hierarchical regression analysis yielded a critical insight: while general attitudes toward digital storytelling did not predict competence (p > 0.05), pedagogical beliefs regarding the use of children’s literature in mathematics were a strong predictor of technical proficiency (β = 0.35, p < 0.001). Qualitative evidence corroborated that students overcame technical limitations not through technological affinity but through a motivation to concretize abstract mathematical concepts via storytelling. These findings suggest that to foster sustainable STEAM education, teacher training curricula must prioritize the “why” (pedagogical conviction) over the “how” (technical mechanics), thereby closing the gap between digital intention and action. This study uniquely demonstrates that domain-specific pedagogical convictions, rather than general technological enthusiasm, are the fundamental drivers of digital competence in STEAM, providing an empirical basis for more resilient teacher education models. Full article

Communicating heritage to contemporary audiences, especially younger generations, is a central challenge in Heritage Education. This study aims to analyse audience perceptions of the documentary Amato Lusitano: Ciência e Humanismo—a cultural mediation tool designed and produced by the authors—and to verify the association of these perceptions with sociodemographic profiles. Framed within an action research methodology, the study combines artistic creation and empirical analysis of public reception. During the first screening, a questionnaire was administered to 55 viewers (37 students aged < 40), and data were submitted to an exploratory factor analysis. This revealed a two-factor structure, highlighting a primary factor—Communicative and Educational Effectiveness—with high internal consistency (α = 0.89). Participants reported high agreement with this dimension, with statistically significant differences across generational groups (p < 0.005). Participants with less prior knowledge about Amato Lusitano showed a greater perception of novelty. Clarity of information and quality of the content were the most valued aspects. In contrast, duration and accessibility emerged as areas for improvement. The potential of the documentary as a heritage education resource for initial teacher training is evident. It is necessary to develop more modular, dynamic content ecosystems tailored for young audiences. Full article