Search Results (597)

International students in Saudi universities face academic and adaptation challenges shaped by emotional, cognitive, linguistic, and sociocultural factors. This study examined whether emotional intelligence and cognitive flexibility predicted academic success and adaptation outcomes among international students in Saudi public universities. A quantitative cross-sectional survey was conducted with 410 international students using structured measures of emotional intelligence, cognitive flexibility, academic success, adaptation outcomes, Arabic proficiency, and sociodemographic characteristics. Data were analysed using descriptive statistics, chi-square tests, Kendall’s tau-b correlations, hierarchical regression, and observed-variable path analysis. Duration of residence was significantly associated with Arabic proficiency, χ²(8) = 82.40, p < .001. Arabic proficiency was positively associated with GPA, τ = 0.62, p < .001, and adaptation outcomes, τ = 0.48, p < .001. In hierarchical regression, sociocultural covariates and psychological predictors explained substantial variance in academic success, R² = 0.53, and adaptation outcomes, R² = 0.53. Emotional intelligence and cognitive flexibility remained positive predictors of both outcomes after accounting for Arabic proficiency, duration of residence, region of origin, and language of instruction. Findings suggest that international student success in Saudi universities reflects an interaction of emotional, cognitive, linguistic, and contextual resources. Universities should strengthen integrated support for emotional regulation, adaptive thinking, Arabic-language development, and culturally responsive academic guidance. Full article

Background: Proficient second language (L2) reading relies on complex neurocognitive processes. Neuroimaging studies have identified key brain regions recruited during L2 reading, including the left inferior parietal lobule (LIPL) and the calcarine cortex (CAL). The LIPL has been suggested to be involved in phonological decoding during L2 reading, whereas the CAL has been implicated in early-stage visual processing. However, given the correlational nature of neuroimaging techniques, it remains unclear whether these regions play causal roles in L2 reading or are merely epiphenomenal. Methods: To address this issue, the present study used transcranial magnetic stimulation (TMS) to modulate neural activity in these regions and eye-tracking technology to assess subsequent reading performance in Chinese–English bilinguals. Specifically, ninety-seven participants were randomly assigned to one of three offline TMS conditions: LIPL, CAL or vertex (as a control site) stimulation, after which they performed a natural sentence reading task in English. Results: The results showed that, compared to the control condition, TMS over the LIPL significantly reduced first fixation duration, whereas no significant effects emerged on gaze duration, regression path reading time, or total reading time. TMS over the CAL produced no significant effects on any eye-movement measures. Conclusions: These findings suggest that the LIPL plays a causal role in L2 reading for early-stage lexical processing through phonological decoding. Overall, this study is the first to employ TMS and eye-tracking to investigate the neural mechanisms underlying natural L2 reading. Full article

The present study investigated the relationship between resilience, self-esteem, and academic performance in primary school students. The sample comprised 124 pupils (59 males and 65 females) enrolled in the 5th and 6th grades. Psychometric assessment was conducted using the Resilience Scale and the Rosenberg Self-Esteem Scale, while academic achievement was evaluated based on students’ grades in Language, Mathematics, Science, English, and Physical Education. Data analysis was performed using ANOVAs and Pearson correlation coefficients. The results indicated that higher academic performance was positively correlated with both increased resilience and self-esteem. Furthermore, a strong positive correlation was observed between self-esteem and resilience. Regarding gender, no significant differences were found in resilience or self-esteem levels. However, academic performance variations were identified exclusively in English language proficiency; specifically, for male students, higher performance in English was significantly associated with greater resilience. In conclusion, these findings suggest that integrating self-esteem and resilience-building activities into the educational curriculum is essential. Such interventions may enhance students’ capacity to manage adversity and facilitate the attainment of their academic goals. Full article

This work presents an innovative approach to test the Arabic language proficiency assessment via Automatic Speech Recognition (ASR) by enhancing the proficiency of the Whisper model in transcribing Arabic speech. The core of our research involved fine-tuning the Whisper model using a substantial, large-scale Arabic speech corpus, with a specific focus on Modern Standard Arabic. This process used a 2000-h Arabic-labeled speech corpus, the QASR dataset, and improved the model’s Word Error Rate (WER). After optimization, the fine-tuned Whisper model’s WER was reduced from 35% to 7% on the QASR dataset, corresponding to an absolute reduction of 28 percentage points (approximately 80% relative reduction). These results demonstrate the strong generalization ability of the fine-tuned model across multiple Arabic ASR benchmarks. A key component of our methodology was the development of a sophisticated scoring system. This system integrates various similarity metrics, such as cosine similarity, the Jaccard index, and the Levenshtein distance, with a machine learning regression model. This multifaceted system provides a comprehensive assessment of reading proficiency, proposing a practical automated assessment method that contributes to the field of AI language transcription and to its application in the assessment of students’ reading. Our research also introduces the ICONET dataset, an augmented Arabic speech corpus comprising 3160 h of diverse and tailored audio–text pairs designed for fine-tuning ASR models. This study demonstrates the potential of fine-tuning pretrained models for specific linguistic contexts (Arabic), establishing a foundation for future research in ASR and language technology. Full article

This study investigates language transmission in immigrant families. The study is based on a bidimensional acculturation model, which assumes that immigrants acquire the new culture and preserve their culture of origin to different degrees. The model was tested using a stratified sample of first-generation immigrants from the former Soviet Union in Israel whose children were born in the host country (n = 725). The assimilation pattern was observed across all components of language transmission, with Hebrew being more prevalent than Russian among parents and children, as well as in their interactions. In addition, the two languages were competitive (negatively correlated) with respect to parents’ language proficiency and parent–child interactions. However, they were complementary (non-correlated) with respect to children’s language proficiency. The hypothesized bidimensional model linking parents’ language proficiency, the frequency of parent–child interactions in a specific language, and children’s language proficiency was corroborated for both languages. In addition, positive effects of parents’ proficiency in Russian on children’s proficiency in both Russian and Hebrew were found. Finally, the duration of residence in Israel, religiosity, education, and gender affected various aspects of language transmission in immigrant families. The study’s results advance our understanding of immigrants’ language acculturation and chart new directions for language policy and practice. Full article

Background/Objectives: This study aimed to investigate how the disparities in unmet medical needs between immigrants to South Korea and native-born populations evolved during the COVID-19 pandemic. Methods: Using nationally representative cross-sectional data from the 2018 and 2020 Surveys on Immigrants’ Living Conditions and Labor Force in South Korea, we compared unmet medical needs among immigrants at two time points (N = 12,227 in 2018; N = 18,530 in 2020). Standardized prevalence ratios (SPRs) were calculated. Analyses were stratified according to work status, gender, Korean language proficiency, education level, and duration of stay. Results: Working immigrants had lower SPRs for unmet medical needs than Korean nationals (2018: 0.879; 2020: 0.745) but non-workers had consistently higher SPRs (2018: 1.117; 2020: 1.128). The SPRs for male and female non-workers increased and decreased, respectively. The SPRs were persistently higher among individuals with poorer Korean language proficiency, lower education, and shorter duration of stay. Conclusions: Systemic disruptions, such as the COVID-19 pandemic, may exacerbate pre-existing healthcare inequalities among immigrant populations. The persistence and widening of these disparities call for targeted policies that address structural barriers and ensure equitable healthcare access during future public health crises. Full article

The design of stretchable energy harvesting systems entails complex multiphysics coupling between electromagnetic and mechanical domains, typically requiring engineers to proficiently use disparate simulation tools and optimization algorithms. This steep learning curve, combined with the absence of integrated workflows, poses a substantial obstacle to efficient design. To overcome these challenges, we present StretchCopilot, a multi-agent collaborative framework driven by Large Language Models (LLMs) for the generative design of stretchable radio frequency (RF) energy harvesters operating in the 2.45 GHz band. In contrast to conventional approaches dependent on manual iteration or isolated algorithmic methods, our framework utilizes a graph-based state machine architecture (LangGraph) to coordinate specialized agents. It interprets high-level user instructions, such as “design a robust energy harvester capable of withstanding 15% strain”, and autonomously manages domain-specific solvers, including inverse design networks and rectifier circuit synthesis tools, through a unified interface. Experimental evaluations indicate that the framework effectively streamlines the design workflow, allowing users to produce desired rectenna (rectifying antenna) systems via natural language interactions. Case studies confirm that, once the underlying surrogate models are fully trained, the proposed approach compresses the marginal design time from several hours to within minutes, while ensuring consistent energy harvesting performance under mechanical deformation. Full article

This study presents a multimodal deep learning framework for automatic proficiency and style classification of parallel Bilingual Tamil–Hindi learner data. The proposed system employs a dual-headed neural architecture to simultaneously predict proficiency levels (Basic, Advanced) and stylistic categories (Formal, Literary) using shared feature representations. A curated dataset of bilingual text samples is utilized, along with synthetic speech generated through text-to-speech (TTS) to enable controlled multimodal experimentation. Five deep learning architectures are evaluated under text-only, audio-only, and learnable fusion settings. Experimental findings indicate that text-based models consistently achieve strong performance in both proficiency and style classification tasks. In contrast, the audio-only model demonstrates limited effectiveness, highlighting the constraints of synthetic acoustic features in capturing meaningful linguistic information. The fusion models provide only marginal improvements over text-based approaches, suggesting that textual representations play a dominant role in proficiency and stylistic classification within controlled datasets. These results emphasize the importance of linguistic features over acoustic signals for automated language assessment in low-resource settings. The proposed framework provides a scalable and reproducible approach and offers a foundation for future work incorporating real speech data and more diverse linguistic inputs. Full article

In this quantitative study, we investigated the conditional mediating role of students’ mathematical language self-efficacy in the relationship between self-regulation and mathematics anxiety. The study employed a relational research design and included survey data from 706 middle school students attending public schools in Turkey. Findings indicated that both self-regulation and perceived self-efficacy in mathematical language use were significantly associated with mathematics anxiety. Moreover, the effect of self-regulation on mathematics anxiety was significantly mediated by students’ perceptions of their ability to understand and use mathematical language self-efficacy. The indirect effect was negative while the direct effect was positive, indicating a suppression (competitive mediation) effect, whereby self-regulation exerts both anxiety-reducing and potentially anxiety-inducing influences through different pathways. Conditional mediation analysis further revealed that this mediating effect varied as a function of students’ perceived academic support, with the indirect effect being non-significant for students who did not receive support. Measurement invariance across gender and grade level was examined to ensure that the constructs were measured equivalently across groups. These findings highlight the importance of fostering both self-regulation skills and mathematical language proficiency, particularly in contexts where students may lack sufficient support. These findings provide a theoretically grounded and practically relevant framework for understanding mathematics anxiety within inclusive mathematics education contexts. Full article

This study investigated the adaptation of culinary culture and behavioral adjustment to Turkish cuisine among international students. The sample comprised 82 students (61.0% males; 39.0% females) from over 20 countries across Europe, Central Asia, South/Southeast Asia, Africa, and the Middle East, all enrolled at Erciyes University. Data collection involved a sociodemographic questionnaire, assessments of food consumption frequency and cooking methods, and the Culinary Culture Adaptation Assessment Inventory. Results indicate that adaptation to Turkish cuisine occurs through a selective and gradual behavioral process. Higher adaptation levels were observed for basic dietary components (bread, soup, rice, yoghurt, and tea), whereas adoption of starch- and sugar-heavy dietary patterns was more limited. Gender comparisons revealed significantly higher scores for meat-heavy and starch-heavy dietary patterns among males (p = 0.048 and p = 0.031, respectively). In contrast, regional origin, economic status, and language proficiency were not significantly associated with culinary acculturation levels. Comparisons based on length of residence identified significant differences in meat-heavy and starch-heavy dietary patterns (p = 0.034 and p = 0.008, respectively). Cooking behaviors remained stable for boiling, grilling, and baking, while frying and roasting decreased. Reported changes in portion perception and body weight suggest that culinary acculturation may extend beyond food choice to broader eating behaviors. Based on these results, an AI-supported interactive platform was developed to facilitate culturally comparable food matching between Turkish and global cuisines. These findings may inform the development of culturally sensitive strategies to support culinary adaptation among international students. Full article

Background: Multimodal large language models (MLLMs) show promise in medical imaging, but their performance is highly dependent on prompt engineering. This study systematically evaluates how different prompting strategies affect diagnostic accuracy in clinical laboratory image interpretation. Methods: We evaluated five MLLMs (ChatGPT-4o, Gemini 2.0 Flash, Claude 3.5 Sonnet, Grok-2, and Perplexity Pro (Claude 3.5 Sonnet)) using 177 proficiency testing images across three domains: blood smears (n = 78), urinalysis (n = 50), and parasitology (n = 49). Three prompting approaches were compared: (1) complex multi-choice prompts with 20 diagnostic options, (2) zero-shot open-ended prompts, and (3) two-step descriptive-reasoning prompts. Images were sourced from the Taiwan Society of Laboratory Medicine external quality assurance archives with expert consensus diagnoses. Results: Zero-shot prompting significantly outperformed complex multi-choice prompts across all models and domains (p < 0.001). With zero-shot prompts, Gemini achieved 78.5% overall accuracy (urinalysis: 92.0%; parasitology: 75.5%; blood smears: 64.1%), representing a 17% improvement over complex prompts. Two-step descriptive-reasoning prompts further improved blood smear accuracy by 8–12% for top-performing models, but showed minimal benefit in urinalysis and parasitology. The re-query mechanism (“please reconsider”) improved urinalysis accuracy by 7.6% but had a negligible effect on blood smears and parasitology. Conclusions: Prompting strategy critically determines MLLM diagnostic performance. Zero-shot approaches with minimal constraints consistently outperform complex multi-choice formats. The remarkable performance of general-purpose models in structured domains like urinalysis (>90% accuracy) demonstrates the considerable progress of multimodal AI. However, complex morphological tasks like blood smear interpretation require either specialized prompting techniques or domain-specific fine-tuning. These findings provide evidence-based guidance for optimizing AI integration in clinical laboratories. Full article

Teachers’ readiness in bilingual early childhood education is increasingly recognized as a multidimensional construct shaped by both professional and language-related factors. However, existing research has typically examined these factors separately, with limited evidence on how they combine across teacher groups, particularly in minority-language contexts. This study examined teachers’ readiness to deliver state-language instruction to dual language learners (DLLs) in Hungarian-medium kindergartens in Slovakia. A total of 313 kindergarten teachers participated in the study. Data were collected through a survey assessing multiple dimensions of readiness. Principal component analysis and confirmatory factor analysis supported a six-factor model comprising professional preparation, teacher competencies, challenge management, instructional aids use, professional needs, and Slovak language use outside kindergarten. Latent profile analysis identified three readiness profiles (low, moderate, and high), reflecting differences in overall preparedness. Background characteristics, particularly age, teaching experience, and language-related factors, were significantly associated with higher readiness. Teachers who used Slovak more frequently in everyday contexts showed higher readiness. Mediation analysis indicated that language proficiency and preferred language use did not mediate the relationship between teaching experience and teachers’ readiness, but functioned as independent predictors. These findings highlight the joint importance of professional and language-related factors in shaping teachers’ readiness and offer implications for teacher education and policy in bilingual early childhood settings. Full article

The rapid expansion of English major enrollment has exposed critical limitations in traditional academic assessment methods regarding efficiency and accuracy, constraining educational quality enhancement. This paper introduces an English proficiency assessment approach utilizing an improved RegNet architecture integrated with a dual attention mechanism. The multidimensional academic data processed by our model include attendance, online participation, language practice, and assessment scores for listening, speaking, reading, and writing from undergraduate English majors. The initial downsampling module of RegNet is optimized through a dual convolutional structure to augment shallow feature extraction. Subsequently, a deformable attention mechanism (DAT) is incorporated to enhance focus on salient features, while a graph attention network (GAT) facilitates interaction and fusion among academic node features. Experimental results demonstrate that the proposed method achieves an average accuracy of 99.46% in proficiency assessment, substantially outperforming mainstream models including EfficientNet and AlexNet. Additionally, it demonstrates robust edge deployment capabilities, providing an effective technical solution for intelligent academic management of English programs within smart campus frameworks. Full article

The rapid advancements in large language models (LLMs) have transformed natural language processing, enabling their application in diverse domains such as conversational agents and decision-support systems in sensitive areas like healthcare, finance, and eldercare. However, as LLMs are increasingly integrated into real-world contexts, concerns about their adherence to ethical principles, privacy norms, and contextual expectations have become critical. Privacy preservation is particularly pressing in interactions involving personal or sensitive data, where ensuring that LLMs align with societal norms while mitigating risks of information leakage is essential to fostering trust and ensuring responsible deployment. Contextual integrity (CI) provides a robust framework to address these challenges, emphasizing that information flows should adhere to context-specific social norms. This principle is especially vital in sensitive applications, where LLMs must evaluate roles, information attributes, and transmission principles to maintain ethical behavior. Despite their linguistic proficiency, LLMs often fail to recognize and adapt to nuanced contextual norms, a limitation exacerbated by their probabilistic nature and the biases in their training data, which can lead to inappropriate or harmful outputs. Addressing these shortcomings requires rigorous evaluation methodologies and fine-tuning strategies that embed societal and contextual norms into the models. Full article

Individual differences in second-language (L2) proficiency are expected to influence how listeners parse and represent continuous speech, yet their neural signatures under naturalistic conditions remain unclear. We investigated this question using task-based fMRI during continuous speech listening. A total of 43 healthy participants completed four listening runs synchronized with MRI acquisition via PsychoPy(Peirce 2007), with eyes open throughout scanning. To promote sustained attention and comprehension, participants provided a native-language oral recall after each run. Based on behavioral proficiency scores, participants were grouped into low- (LP, n = 14), moderate- (MP, n = 14), and high-proficiency (HP, n = 15) groups. We evaluated three temporal information-encoding frameworks derived from BOLD dynamics: direct temporal series, functional connectivity (FC), and self-information weighted inter-subject correlation (ISC-W). Using a 10 × 5-fold nested cross-validation scheme, we tested both categorical classification (Support Vector Machines) for discrete proficiency groups (LP, MP, HP) and continuous multivariate regression (Ridge/Lasso) for continuous proficiency scores. Furthermore, we applied ROI-based ANOVA and univariate Neural Correlation Analysis (NCA) to identify key brain regions, evaluating significance via nonparametric permutation testing (1000 permutations) and False Discovery Rate (FDR) correction. Results indicated that while categorical classification yielded numerical trends—with ISC-W performing best—it did not reach statistical significance under stringent permutation testing. However, multivariate continuous regression using ISC-W features successfully predicted continuous proficiency scores with statistical significance (p < 0.05). Exploratory ROI analysis highlighted the bilateral orbital inferior frontal gyrus (IFG_orb_bilat) as a highly sensitive region. These findings suggest that L2 proficiency is best represented as a distributed, continuous neural variable, and that self-information weighting effectively filters background noise to capture cognitive variance. Methodologically, this study provides a reproducible pipeline integrating information-theoretic feature construction with rigorous whole-brain nonparametric inference. Full article