Search Results (3,030)

This study evaluated the accuracy and educational potential of three generative AI models, ChatGPT 3.5, ChatGPT 4o, and Gemini 2.5, by addressing pharmacy-related content across three key areas: biostatistics, pharmaceutical calculations, and therapeutics. A total of 120 exam-style questions, categorized by Bloom’s Taxonomy levels (Remember, Understand, Apply, and Analyze), were administered to each model. Overall, the AI models achieved a combined accuracy rate of 77.5%, with ChatGPT 4o consistently outperforming ChatGPT 3.5 and Gemini 2.5. The highest accuracy was observed in therapeutics (83.3%), followed by biostatistics (81.7%) and calculations (67.5%). Performance was strongest at lower Bloom levels, reflecting proficiency in recall and conceptual understanding, but declined at higher levels requiring analytical reasoning. These findings suggest that generative AI tools can serve as effective supplementary aids for pharmacy education, particularly for conceptual learning and review. However, their limitations in quantitative and higher-order reasoning highlight the need for guided use and faculty oversight. Future research should expand to additional subject areas and assess longitudinal learning outcomes to better understand AI’s role in improving critical thinking and professional competence among pharmacy students. Full article

As education undergoes digital transformation, ChatGPT-4 has emerged as one of the most visible tools of generative artificial intelligence. While widely discussed, its impact on student satisfaction and learning outcomes in higher education remains underexplored. This study investigates the factors that shape art and design students’ satisfaction when using ChatGPT to support coursework. Unlike previous research focusing on ChatGPT adoption behavior, this study extends the Information Systems Success Model (ISSM) to the context of art and design education. Drawing on 435 valid survey responses, we employed a mixed-methods approach. Partial Least Squares Structural Equation Modeling (PLS-SEM) was first applied to examine how system quality, compatibility, personal innovativeness, and perceived usefulness influence satisfaction directly and through mediating mechanisms. To complement this, fuzzy-set Qualitative Comparative Analysis (fsQCA) was used to identify multiple combinations of conditions that lead to high satisfaction. The findings show that compatibility, perceived usefulness, and personal innovativeness significantly enhance satisfaction, with path coefficients of 0.378, 0.342, and 0.155, respectively. Importance–Performance Map Analysis (IPMA) further highlights personal innovativeness and system quality as critical drivers. By providing both theoretical and practical insights, this study contributes to the growing body of research on generative AI in art and design education and informs the design of courses and digital learning tools. Full article

Advancements in data storage and data processing technologies has compelled higher education institutions to optimise the use of their data. Many universities globally have begun to implement learning analytics at their institutions to better understand and improve teaching and learning. African higher education institutions have been slow to implement learning analytics despite the continued accumulation of digital data. The research related to this study presents a dataset of Information Systems and Technology (IS&T) students from the University of KwaZulu-Natal, a South African university. The dataset comprises approximately 14,000 registered student records from 10 IS&T courses, primarily consisting of demographic data, academic performance (including past IS&T courses and school records), and Learning Management System (LMS) interaction data. The dataset exhibits an imbalance, characterised by a higher proportion of students who have successfully completed courses compared to those who have not. The dataset will be of interest to researchers engaged in learning analytics application studies, including early pass/fail prediction and grade classification, as well as those who want to test their techniques on a real-world dataset. Full article

Semi-supervised learning (SSL) is critical for medical image segmentation but often struggles with network dependency and pseudo-label error accumulation. To address these issues, we propose a fractal-dimension-guided hierarchical contrastive learning dual-student framework(FD-HCL). We extend the Mean Teacher architecture with a dual-student design and introduce an independence-aware exponential moving average (I-EMA) update mechanism to mitigate model coupling. For enhanced feature learning, we devise a hierarchical contrastive learning (HCL) mechanism guided by voxel uncertainty, spanning global, high-confidence, and low-confidence regions. We further improve structural integrity by incorporating a fractal-dimension (FD)-weighted consistency loss and integrating a novel uncertainty-aware bidirectional copy–paste (UB-CP) augmentation. Extensive experiments on the LA and BraTS 2019 datasets demonstrate the state-of-the-art performance of our framework across 10% and 20% labeled data settings. On the LA dataset with 10% labeled data, our method achieved a Dice score that outperformed the best existing approach by 0.68%. Similarly, under the 10% labeling setting on the BraTS 2019 dataset, we surpassed the state-of-the-art Dice score by 0.55%. Full article

Early identification of students at risk for low academic confidence, poor problem-solving skills, or poor academic performance is crucial to achieving equitable and sustainable learning outcomes. This research presents a hybrid artificial intelligence (AI) framework that combines feature selection using a Random Forest (RF) algorithm with data classification via an Artificial Neural Network (ANN) to predict risks related to Academic Confidence and Problem-Solving Skills (ACPS) among higher education students. Three real-world datasets from Saudi universities were used: MSAP, EAAAM, and MES. Data preprocessing included Min–Max normalisation, class balancing using SMOTE (Synthetic Minority Oversampling Technique), and recursive feature elimination. Model performance was evaluated using five-fold cross-validation and a paired t-test. The proposed model (RF-ANN) achieved an average accuracy of 98.02%, outperforming benchmark models such as XGBoost, TabNet, and an Autoencoder–ANN. Statistical tests confirmed the significant performance improvement (p < 0.05; Cohen’s d = 1.1–2.7). Feature importance and explainability analysis using a Random Forest and Shapley Additive Explanations (SHAP) showed that psychological and behavioural factors—particularly study hours, academic engagement, and stress indicators—were the most influential drivers of ACPS risk. Hence, the findings demonstrate that the proposed framework combines high predictive accuracy with interpretability, computational efficiency, and scalability. Practically, the model supports Sustainable Development Goal 4 (Quality Education) by enabling early, transparent identification of at-risk students, thereby empowering educators and academic advisors to deliver timely, targeted, and data-driven interventions. Full article

Personalized e-learning emphasizes a learner-centered instruction approach by adapting educational content to individual needs. This study empirically evaluates three personalization methods (adaptive, adaptable, and mixed) alongside a non-personalized static method, implemented within Moalemy, an Arabic rule-based educational chatbot. A 4 × 3 within-subjects experiment involving 52 students across three levels of task difficulty (easy, medium, and hard) was performed to examine usability, learner engagement, and learning outcomes as key indicators of user experience in educational chatbots. Results showed no statistically significant differences in learning gain or relative learning gain across the four methods, indicating comparable effectiveness in supporting learning success. Engagement differed significantly by method, with the adaptable approach obtaining the highest scores. Descriptive usability results further suggested that the adaptable method achieved numerically higher scores in SUS, efficiency, and effectiveness under lower task difficulty, while static and adaptive showed comparatively stronger tendencies at higher levels. However, these patterns represent exploratory trends rather than statistically confirmed performance advantages. This study provides a controlled comparison of system-driven, learner-driven, shared-control, and non-personalization strategies in an Arabic e-learning context, offering empirical insights into how different learning methods influence usability, learning success, and engagement within Arabic educational chatbots. Full article

This study aimed to design a virtual reality (VR)–based mountaineering education system and examined its effects on junior high school students’ learning outcomes, motivation, and cognitive load in compass operation and map reading. The system integrated 3D terrain models and interactive mechanisms across four instructional modules: Direction Recognition, Map Symbols, Magnetic Declination Adjustment, and Resection Positioning. By incorporating immersive 3D environments and hands-on virtual exercises, the system simulates authentic mountaineering scenarios, enabling students to develop essential field orientation and navigation skills. An experimental design was implemented, with participants assigned to either an experimental group learning with the VR system or a control group receiving slide-based instruction. Data were collected using pre-tests, post-tests, and questionnaires, and analyzed using SPSS for descriptive statistics, paired-sample t-tests, independent-sample t-tests, and one-way ANCOVA at a significance level of α = 0.05. The findings indicated that the experimental group achieved significantly higher post-test learning performance than the control group (F = 6.37, p = 0.014). Moreover, significant or highly significant improvements were observed across the four dimensions of learning motivation—attention, relevance, confidence, and satisfaction. The experimental group also exhibited a significantly lower extraneous cognitive load (p = 0.024). Therefore, the VR mountaineering education system provides an immersive, safe, and effective approach to teaching mountaineering and outdoor survival skills. Full article

Technology plays an increasingly vital role in modern education, providing new opportunities to enhance engagement and conceptual understanding. Among emerging innovations, Augmented Reality (AR) enables interactive visualization that supports deeper comprehension of abstract and spatially complex concepts. This study aimed to evaluate the impact of AR technology integrated with the STEAM approach on fifth-grade students’ learning of geometric solids, focusing on spatial skills, motivation, and academic achievement. A quasi-experimental design was implemented, involving an experimental group that engaged in AR- and STEAM-based activities and a control group that followed traditional instruction. Results indicated significant improvement in geometry test performance within the experimental group (p < 0.001) and higher post-test performance compared to the control group (p = 0.005). Although motivation scores were higher in the experimental group, the difference was not statistically significant (p = 0.083), suggesting a positive trend that merits further exploration with a larger sample. Overall, the findings highlight the pedagogical potential of integrating AR and STEAM approaches to support engagement and conceptual understanding in geometry education. Full article

Self-efficacy beliefs and mindset influence student success, impacting how a learner experiences and responds to learning situations and setbacks. Accordingly, mindset interventions, are successful at increasing student performance with particular efficacy with historically underserved students such as those attending HBCUs. This paper studies a classroom-based mindset intervention that was implemented with the goal of increasing learning and achievement through improving the students’ cognitive disposition. The intervention, implemented at a mid-Atlantic minority serving institution of higher education, involved the creation of a custom-designed three-tool self-assessment developed to engender students’ critical reflection. The scales in question measured self-efficacy, growth mindset, and mastery goal orientation. This paper presents the results of reliability testing via Cronbach’s alpha and inter-item covariance. According to the findings, all three tools showed strong (good to excellent) reliability with acceptable positive covariance indicating that they are capable of serving as appropriate instruments for further adoption, usage, and analysis. It is the goal that this paper contributes to the body of literature on mindset interventions encouraging more individuals working with traditionally underserved learners to consider exploring efforts to increase students’ positive mindsets. Full article

Screencasts, which are screen-capture videos, have been created by teachers delivering instruction or feedback, reflecting a teacher-centered model of learning. Based on the constructivist principle, this study explores an innovative attempt to position students as screencast creators, who must demonstrate their knowledge by and explain their work in the screencast. This innovative approach has the potential to promote authentic learning and reduce dependence on generative artificial intelligence (GenAI) tools for completing assignments. However, it is uncertain whether students will have positive attitudes towards this new form of assessment. From 2022 to 2025, the authors used screencasts as assessments in computer programming and English language subjects. Survey results were obtained from 203 university students and analyzed using Partial Least Squares Structural Equation Modeling (PLS-SEM). The results show that students generally hold positive attitudes toward creating screencasts, with perceived usefulness for future applications exerting the strongest influence on acceptance, followed by perceived performance benefits and ease of use. It is also found that gender, discipline, and study mode did not significantly alter these relationships, although senior students perceived screencast production as more effortful. These findings suggest that student-created screencasts can serve as an effective, student-centered alternative to traditional written assessments. The research results imply that student-created screencasts have the potential to help students develop their skills in an increasingly GenAI-pervasive academic environment. Full article

Gender mainstreaming in social work education requires moving beyond policy commitments to ensure that gender perspectives are meaningfully integrated into teaching and learning. This study examines how gender-responsive pedagogy is implemented in a Spanish public university and how these practices relate to students’ self-efficacy for gender-sensitive social work. A sample of 166 undergraduate students completed validated measures of gender-responsive teaching, self-efficacy, and institutional climate. The instruments demonstrated strong psychometric performance. Results indicate that while gender-related content is incorporated into curricula, practice-oriented and participatory pedagogies are less consistently used. Students reported high confidence in gender knowledge and attitudes but lower confidence in applied skills. Teaching methods, rather than content coverage, showed the strongest associations with self-efficacy. Institutional reforms at the degree and course levels were positively linked to teaching practices and student outcomes, whereas governance-level changes showed weaker associations. These findings highlight the importance of aligning institutional commitments with pedagogical innovation to advance gender equality in social work education. Full article

Educational chatbots are increasingly deployed to scaffold student learning, yet educators lack scalable ways to assess the cognitive depth of these dialogues in situ. Bloom’s taxonomy provides a principled lens for characterizing reasoning, but manual tagging of conversational turns is costly and difficult to scale for learning analytics. We present a reproducible high-confidence pseudo-labeling pipeline for multi-label Bloom classification of Socratic student–chatbot exchanges. The dataset comprises 6716 utterances collected from conversations between a Socratic chatbot and 34 undergraduate statistics students at Nanyang Technological University. From three chronologically selected workbooks with expert Bloom annotations, we trained and compared two labeling tracks: (i) a calibrated classical approach using SentenceTransformer (all-MiniLM-L6-v2) embeddings with one-vs-rest Logistic Regression, Linear SVM, XGBoost, and MLP, followed by per-class precision–recall threshold tuning; and (ii) a lightweight LLM track using GPT-4o-mini after supervised fine-tuning. Class-specific thresholds tuned on 5-fold cross-validation were then applied in a single pass to assign high-confidence pseudo-labels to the remaining unlabeled exchanges, avoiding feedback-loop confirmation bias. Fine-tuned GPT-4o-mini achieved the highest prevalence-weighted performance (micro-F1 $=0.814$), whereas calibrated classical models yielded stronger balance across Bloom levels (best macro-F1 $=0.630$ with Linear SVM; best classical micro-F1 $=0.759$ with Logistic Regression). Both model families reflect the corpus skew toward lower-order cognition, with LLMs excelling on common patterns and linear models better preserving rarer higher-order labels, while results should be interpreted as a proof-of-concept given limited gold labeling, the approach substantially reduces annotation burden and provides a practical pathway for Bloom-aware learning analytics and future real-time adaptive chatbot support. Full article

This study investigates the acoustic conditions of a design studio (Studio 130) in the Department of Interior Architecture and Environmental Design at Afyon Kocatepe University by integrating 14 weeks of continuous noise measurements with perception data collected from 192 students. Noise measurements were conducted in accordance with ISO 3382-3:2022 guidelines at three locations—window front, door side, and studio midpoint—during morning, noon, and evening periods, with 10 min recordings at each session. The results indicate that when students were present, the equivalent continuous noise level (Leq) reached an average of 65.5 dB(A), with peak levels rising to 72.3 dB(A) during jury sessions. These values substantially exceed the recommended 35 dB(A) classroom threshold by the World Health Organization and the 35–45 dB(A) limits specified in national regulations for indoor educational spaces. Survey findings reveal that 88% of students experienced loss of concentration, 72% reported decreased productivity, 60% had difficulty communicating, and 52% reported fatigue due to noise exposure. Pearson correlation analysis demonstrated a strong relationship between measured noise levels and reported negative effects (r = 0.966). Moreover, independent samples t-test results confirmed that student presence significantly increased studio noise levels (t = 4.98, p < 0.001). The novelty of this research lies in its combined use of longitudinal objective measurements and subjective perception data, addressing the unique open-plan, collaborative, and critique-based pedagogical structure of design studios. The findings highlight that acoustic comfort is a critical component of learning quality in studio-based education. Based on the results, the study proposes several design and material interventions—including spatial dividers, acoustic ceiling panels, fabric-wrapped absorbers, and impact-reducing flooring—to enhance auditory comfort. Overall, the study emphasizes the necessity of integrating acoustic design strategies into studio pedagogy to support concentration, communication, and learning performance. Full article

Large language models (LLMs) are rapidly being explored as tools to support learning and assessment in health science education, yet their performance across discipline-specific evaluations remains underexamined. This study evaluated the accuracy of two prominent LLMs on university-level pharmacognosy examinations and compared their performance to that of pharmacy students. Authentic exam papers comprising a range of question formats and content categories were administered to ChatGPT and DeepSeek using a structured prompting approach. Student data were anonymized and LLM responses were graded using the same marking criteria applied to student cohorts, and a Monte Carlo simulation was conducted to determine whether observed performance differences were statistically meaningful. Facility Index (FI) values were calculated to contextualize item difficulty and identify where LLM performance aligned or diverged from student outcomes. The models demonstrated variable accuracy across question types, with a stronger performance in recall-based and definition-style items and comparatively weaker outputs for applied or interpretive questions. Simulated comparisons showed that LLM performance did not uniformly exceed or fall below that of students, indicating dimension-specific strengths and constraints. These findings suggest that while LLM-resistant examination design is contingent on question structure and content, further research should refine their integration into pharmacy education. Full article

Statistical education is a crucial yet often overlooked aspect of AI in higher education. However, traditional approaches usually focus heavily on procedural knowledge, leaving students anxious about statistics and less confident in applying concepts to real-world problems. This study examines a method that enhances statistical learning outcomes by integrating data visualization and gamification strategies. Students were randomly assigned to either a control group (CG) or an intervention group (IG), and each group was further divided into teams. The curriculum was enhanced in a college statistics course offered for both engineering and math majors. IG students applied data visualization and gamification in a hands-on group project aimed at solving a real-world problem and competed as they presented their results. The effectiveness of this approach was assessed through statistical analyses comparing the performance of IG and CG in surveys, final grades, and project grades. The results from evaluation methods indicated that IG students outperformed CG students, demonstrating a positive impact of gamification on statistics education. Full article