Search Results (150)

Objectives: Despite high overall vaccination coverage in Galicia, Spain, human papillomavirus (HPV) vaccine uptake remains below the 90% target set by the World Health Organization for 2030. This study aimed to assess baseline knowledge of HPV and attitudes towards HPV vaccination among Galician adolescents and to evaluate the impact of a brief educational intervention delivered as a “pill of knowledge”. Methods: A quasi-experimental pre-/post-intervention study was conducted among 967 students aged 12–16 years from 16 secondary schools in Galicia during the 2023–2024 academic year. A concise, structured 15-min educational session termed a “pill of knowledge” was delivered, and HPV-related knowledge and vaccination intention were measured immediately before and after the intervention using a standardized questionnaire. Results: Following the “pill of knowledge”, the mean proportion of correct responses increased by 30.1 ± 16.6% across all knowledge items. Among unvaccinated participants, intention to accept HPV vaccination rose from 77.7% to 94.4% in girls and from 64.7% to 85.8% in boys. Pre-intervention predictors of vaccination intention included perceived vaccine efficacy and baseline HPV knowledge. Post-intervention independent predictors comprised being female, younger age (12–13 years), and prior sexual education delivered by teachers or parents. The overall predictive accuracy of the logistic regression model for vaccination intention improved from 75.6% before the intervention to 92.7% afterwards. Conclusions: A brief, school-based “pill of knowledge” produced substantial and immediate improvements in HPV knowledge and vaccination acceptance among Galician adolescents. These findings strongly support the systematic incorporation of short, evidence-based educational interventions of this kind into the school setting as an effective public health measure to increase HPV vaccine coverage and advance progress toward WHO elimination targets. Full article

This paper proposes an Online Prototype Angular Balanced Self-Distillation (OPAB) framework to address the challenges posed by non-ideal annotation in remote sensing image semantic segmentation. “Non-ideal annotation” typically refers to scenarios where long-tailed class distributions and label noise coexist in both training and testing sets. Existing methods often tackle these two issues separately, overlooking the conflict between noisy samples and minority classes as well as the unreliable early stopping caused by non-clean validation sets, which exacerbates the model’s tendency to memorize noisy samples. OPAB mitigates the imbalance problem by employing an improved bilateral-branch network (BBN) that integrates max-min angular regularization (MMA) and category-level inverse weighting to achieve balanced hyperspherical representations. The balanced hyperspherical representations further facilitate noise-clean sample separation and early stopping estimation based on large category-wise Local Intrinsic Dimensionality (LID). Moreover, OPAB introduces a bootstrap teacher label refinement strategy coupled with a student full-parameter retraining mechanism to avoid memorizing noisy samples. Experimental results on ISPRS datasets demonstrate that OPAB achieves a 2.0% mIoU improvement under non-ideal annotation conditions and achieves 89% mIoU after cross-set correction, showcasing strong robustness across different backbones and effective iterative calibration capability. Full article

Financial report question answering (FRQA) presents unique challenges due to the need for precise numerical reasoning, complex table structures, and multi-table associations. Existing approaches often overlook the domain-specific complexities of financial reports and struggle with accurate numerical computation, leading to suboptimal performance in real-world financial intelligence applications. In this study, we propose FinQA-PKD, a framework designed to mitigate these challenges through a novel integration of progressive knowledge distillation and numerical reasoning enhancement. Our method introduces a difficulty-aware curriculum learning strategy that organizes training into two progressive stages, facilitating more effective and stable model learning. To address the limitations of large language models in numerical reasoning, we develop a numerical reasoning enhancement module that automatically decomposes calculation chains, augments numerical tokens, and validates results using a financial formula library. Furthermore, we implement a domain-adaptive selective knowledge distillation strategy, which evaluates teacher model outputs based on numerical accuracy, calculation correctness, and terminology precision, and selectively distills knowledge from high-quality samples. Experimental results in benchmark datasets demonstrate that FinQA-PKD improves numerical and calculation accuracy, achieving competitive performance with reduced computational resources. This framework provides a robust and efficient solution for answering financial report questions in practical financial analysis scenarios. Full article

Background: Nutrition problems among primary school children increase the risk of illness, reduce school attendance, and impair academic performance. The Indonesian national school feeding program (ProGAS—Program Gizi Anak Sekolah) was developed to address these issues through the provision of healthy breakfast, nutrition education, and character building. Methods: This study employed a mixed-methods design involving 454 primary school students aged 8–14 years from 24 schools across four provinces. Data collection included structured questionnaires, 24 h dietary recalls, and anthropometric measurements. In-depth interviews with school principals, teachers, cooking teams, parents, students, nutritionists, and district education office staff were conducted to capture experiences and opinions on the ProGAS implementation. Results: ProGAS significantly improved students’ dietary diversity, meal frequency, handwashing with soap, and nutrition knowledge, alongside a positive trend in school attendance. Improvements included higher mean dietary diversity, increased proportion of students eating ≥3 meals/day, greater correct responses on balanced nutrition, and higher rates of handwashing before meals (all p < 0.01). Children also developed positive behaviors such as praying before meals, queuing, and taking responsibility for cleaning dishes. However, students’ breakfast habits did not significantly improve, the delivery of nutrition education was suboptimal, and no significant changes were observed in nutrient intakes or nutritional status based on BMI-for-age. While the energy and protein contributions of the ProGAS menu met the recommended 25–30% of daily requirements for breakfast, its micronutrient contributions remained below the recommended levels. Key management gaps include the delivery of nutrition education to students and the monitoring of implementation by local and national authorities. Conclusions: ProGAS demonstrated positive impacts on some dietary and hygiene practices as well as learning environment. To achieve greater improvements in breakfast habits, nutrient intake and nutritional status, it is recommended to strengthen the school feeding menus not only for dietary diversity but also for nutrient density, enhance capacity building for teachers, deliver regular and engaging nutrition education, and reinforce program monitoring. Full article

Graph anomaly detection (GAD) aims to identify nodes or edges that deviate from normal patterns. However, the presence of heterophilic edges in graphs leads to feature over-smoothing issues. To overcome this limitation, this paper proposes the multi-view heterogeneity resistant network (MV-GHRN) model, which progressively purifies heterophilic edges through multi-view collaboration. First, to address the noise sensitivity of single predictions, the method computes post-aggregation (PA) scores for both the original graph and its perturbed versions and performs weighted fusion, leveraging the consistency of multiple prediction perspectives to enhance the reliability of heterophilic edge identification. Second, a cosine similarity view is introduced as a complementary structural perspective, with both views independently completing heterophilic edge pruning to clean the graph structure from both topological and feature dimensions. Finally, a cross-view self-distillation mechanism is designed, using the fused predictions from the two purified views as teacher signals to guide the optimization of each view in reverse, correcting feature biases caused by heterophilic edges. Experiments on benchmark datasets such as YelpChi and Amazon demonstrate that the framework significantly outperforms existing methods. For instance, on the YelpChi dataset, MV-GHRN surpasses the best baseline by 16.8% and 5.2% in F1-Macro and AUC, respectively, validating the effectiveness of the progressive multi-view purification mechanism. Full article

Background/Objectives: More than half of out-of-hospital cardiac arrests occur at home and are witnessed by family members, who must promptly call the Emergency Medical Services (EMS). The aim of this study was to assess the learning outcomes of an interactive school-based training activity focused specifically on the EMS call. Methods: A single-group, post-test-only simulation study was conducted in five Spanish schools. Participating schoolchildren received basic life support (BLS) training from their Physical Education teachers, integrated into the regular school schedule and following the Kids Save Lives recommendations. An innovative didactic resource (the “BLS Endless Book”) was used to support active learning. Children’s performance was evaluated in a simulated scenario using a standardized checklist. Results: A total of 1341 children aged 6 to 14 years participated. In the simulated scenario, more than 90% of participants were able to recognize the emergency and correctly identified and dialed the national emergency number. However, less than 50% were able to activate the hands-free function (with younger children experiencing more difficulty). During the call, 99.0% stated their full name, and 82.0% provided their complete address. Conclusions: A short, focused BLS training led at school by physical education teachers and based on an interactive, easy-to-use, didactic tool is effective in educating 6–14-year-old schoolchildren to correctly perform an immediate EMS call in case of cardiac arrest. Nevertheless, schoolchildren may require reinforcement training focused on hands-free operation and on providing the correct address. Full article

Students with autism spectrum disorder (ASD) often face difficulties in solving arithmetic word problems, particularly in transitioning from informal counting strategies to more efficient methods based on number facts and formal operations. This study examined the development of problem-solving strategies in a single student with ASD and intellectual disability across two sequential single-case experiments using multiple baseline designs. Study 1 (age 13 years 9 months; 17 sessions) employed Modified Schema-Based Instruction (MSBI) to teach addition and subtraction change problems, while Study 2 (age 14 years 10 months; 18 sessions) utilized the Conceptual Model-based Problem Solving (COMPS) approach for multiplication and division equal-group problems. Success was defined as both correctness of the response and correctly identifying the required operation. Results indicated that the student’s performance improved in all problem types in both studies, with maintenance observed 8 weeks after Study 1 and 5 weeks after Study 2. Instruction effects generalized to two-step addition and subtraction problems in Study 1, and to two-step addition and multiplication problems in Study 2. The findings indicate that both MSBI and COMPS facilitated the student’s shift from informal strategies to efficient operation-based problem solving. Implications for practice include the need for individualized reinforcements, careful adaptation of instruction, and providing teachers with a variety of problems and knowledge of these teaching methods to support students with ASD in developing advanced problem-solving skills. Full article

This study evaluates the impact of using group concept maps in the teaching of Applied Thermodynamics in the Bachelor’s Degree in Industrial Electronics and Automation Engineering. The methodology consisted of selecting topics with a high conceptual load, collaboratively creating concept maps, and subsequently evaluating them by both students and teaching staff. Students achieved average scores above 7/10 in the concept map activity, with teacher and student evaluations averaging 7.8 and 7.3, respectively. Knowledge assessment via pre- and post-tests revealed a 20% increase in concept comprehension. For example, in the topic of Principles of Thermodynamics, the percentage of correct answers on the most complex question increased from 13% in the Pre-Test to 40% in the post-test. In the topic of Refrigeration Cycles, some questions showed an improvement from 18% to 25%. The students’ perception of the activity was positive, with an average satisfaction rating of 6.9 out of 10. Furthermore, most students acknowledged that the activity helped them stay engaged with the subject matter and identify errors in their own learning. The high participation in the activity, despite its low impact on the final grade, demonstrates the students’ strong motivation for this study approach. Therefore, the implementation of concept maps not only facilitated the understanding of key concepts but also promoted critical reflection and collaborative learning, establishing itself as an effective strategy in the teaching of Applied Thermodynamics. Full article

Reinforcement Learning from Human Feedback (RLHF) produces powerful instruction-following models but relies on a preference-labeling process that is both costly and slow. An effective alternative, Reinforcement Learning from AI Feedback (RLAIF), uses large language models as teachers for relabeling; however, this introduces substantial label noise. In our setting, we found that AI teachers flipped approximately 50% of the original human preferences on the dataset, a condition that degrades the performance of standard direct preference optimization (DPO). We propose noise-robust DPO (nrDPO) and nrDPO-gated, two drop-in variants that make DPO resilient to noisy preferences. nrDPO reweights each pair by (i) a margin-confidence term from a frozen reference policy (base or SFT), (ii) a context-stability term that penalizes preferences that change under truncated histories, and (iii) a length correction to curb verbosity bias. nrDPO-gated further filters low-confidence pairs via a simple threshold on the reference margin. On a dataset with heavy synthetic noise (30% flips), nrDPO-gated improves the preference accuracy by +3.8% over vanilla DPO; in a realistic RLAIF setting, nrDPO-gated is the only configuration that recovers competitive alignment, reaching ≈60% on a 5k relabeled set (vs. ≈49–50% for vanilla DPO) and approaching RLHF baselines. Full article

The incorporation of educational robotics (ER) into classroom learning has emerged as a significant goal in contemporary education, with instructors assuming a pivotal role. Recent research has shown the influence of teachers’ perceptions of ER and their self-efficacy on the learning process, while the primary goal in these inquiries is to the development of appropriate scales that guarantee correct measurements. Serving this goal, the present study presents the TRi-Robotics scale and its psychometric properties, which assesses teachers’ readiness to integrate ER into their classrooms. TRi-Robotics is a novel multidimensional tool that integrates self-efficacy, commitment, and affective conditions, validated through both CFA and network psychometrics. The proposed 14-item scale is three-dimensional and includes self-efficacy (SE), commitment (C), and affective conditions (AC). The validation procedure included the customary Exploratory and Confirmatory Factor Analysis, applied to a sample of 817 in-service teachers. Reliability analysis showed satisfactory internal consistency, while measurement invariance for gender was sustained. Furthermore, network psychometrics was applied via Exploratory Graph Analysis (EGA), which supported the proposed structure and its dimensionality and measurement invariance. The TRi-Robotics scale proved a valid instrument with satisfactory psychometric properties, and it is a significant asset to implement in educational and psychological research for testing further research hypotheses. Full article

The application of artificial intelligence in education, particularly in learning programming, is gaining increasing significance. However, research on secondary school students specializing in IT at an early stage has received relatively little attention in this field. The aim of this study is to assess how vocational secondary school IT students utilize Generative artificial intelligence in learning programming. The study employed a survey-based methodology, where students with varying levels of knowledge were surveyed to understand their AI usage patterns. The sample consisted of students from vocational IT schools, and data were analyzed using descriptive statistics and independent samples t-tests. The results indicate that students with different levels of knowledge use AI tools differently, with ChatGPT being the most popular tool. The study further highlights that AI usage brings significant benefits, such as providing a personalized learning experience and enabling quick error correction. However, excessive reliance on AI tools may hinder students from acquiring fundamental programming skills. The findings support the idea that while AI can effectively complement teachers’ explanations, overdependence on it can be risky, potentially reducing students’ creativity and problem-solving abilities. The study emphasizes the crucial role of educators in teaching the responsible and ethical use of artificial intelligence. The results of this research offer new perspectives on the effective integration of Generative artificial intelligence into vocational secondary school programming education and suggest further studies to compare its applications at the university level. However, the study acknowledges certain limitations, such as the potential bias of self-reported data, which may affect the generalizability of the results. Unlike other studies, the age groups we surveyed, and the cohorts formed from them are nearly evenly distributed, making our sample representative of the region in question. Full article

The purpose of this research is to investigate the misconceptions related to geological concepts among Secondary Education students in the region of Achaia, Greece. The study focuses on both Lower Secondary Education (Gymnasium, grades 7–9) and Upper Secondary Education, including General and Vocational Education (grades 10–12). Previous research has shown that students entering Lower Secondary Education or High School often possess several misconceptions about geological concepts. These misconceptions result in a fragmented or incorrect understanding, which may arise from intuitive perceptions of how the natural world evolves that are incorrect, or from stereotypes and assumptions acquired from the family environment or inadequacies in the school curriculum. Despite teachers’ efforts to clarify these concepts, a significant percentage of students continue to hold misconceptions, mainly related to minerals and rocks. A total of 1065 secondary students completed an online closed-ended questionnaire that was designed and validated based on previous research findings to highlight their misconceptions. This study results showed a clear differentiation between students from urban and rural areas, while demographic characteristics (such as gender, age, parents’ occupation, and parents’ marital status) did not appear to play a significant role. In addition, the responses to specific sets of questions varied depending on the student’s grade level. Identifying students’ misconceptions can support the development of appropriate educational tools and/or inform targeted interventions that aim to clarify these concepts and correct any incorrect assumptions. Full article

This study investigates the influence of Trait Emotional Intelligence (TEI) on affective dimensions of English language learning among 515 Pakistani EFL learners, addressing a key gap in Global South research. Using bootstrapped multiple regression and culturally adapted instruments (Cronbach’s α = 0.724–0.857), findings reveal that in Pakistan’s exam-driven, teacher-centered classrooms, well-being significantly enhances attitudes (β = 0.172, p < 0.001), motivation (β = 0.219, p = 0.002), and engagement (β = 0.179, p < 0.001). Emotionality, however, increases anxiety (β = 0.192, p < 0.001) and lowers engagement (β = −0.092, p = 0.025), contradicting global models due to punitive error correction. Sociability shows no significant effect (attitudes: β = 0.038, p = 0.366; engagement: β = 0.019, p = 0.613), reflecting limited peer interaction in hierarchical classrooms. Notably, an emergent auxiliary facet—contextual adaptability—strongly predicts motivation (β = 0.269, p < 0.001) and anxiety (β = 0.109, p = 0.020), highlighting the role of competencies like Urdu–English code-switching. These results call for a Contextually Stratified TEI Framework, emphasizing that while well-being is universal, other TEI dimensions are context-dependent. Implications urge educators to foster well-being, reframe emotionality as a risk-detection skill, and promote adaptability to local linguistic realities. Full article