Abstract

Indonesia's education system is dealing with issues worldwide, such as fake diplomas, poor administration, fragmentation of data, and a misalignment of curriculum and job market demands. The centralised data system of the Indonesian government is hackable/ alterable. This study aims to explore and evaluate global blockchain technology's relevance to and challenges in the Indonesian education system. This study applied the Literature Review methodology with the use of the PRISMA framework. Out of the numerous studies available, 49 were chosen from the Scopus database from 2016 to 2026. The results of the study indicate that blockchain technology offers education systems a solution to the problem of storage, from the perspective of security, transparency, and immutability. Furthermore, the problem of diploma fraud and administrative inefficacy in systems can be solved through the decentralised verification of blockchain. Adopting this technology is still a problem, in that educational data remains unregulated and the internet is poorly available; the large storage capacity of the blockchain system is a problem too. Overall, blockchain technology has the possible prospects of transforming the Indonesian education system to be more inclusive, accountable and secure. To achieve this, unified efforts of all stakeholders is essential

Abstract

Variation orders (VOs) represent a persistent source of cost overruns, schedule delays, and stakeholder disputes in construction projects worldwide, yet their differential impact on performance across heterogeneous causal categories remains insufficiently understood, particularly in specialised marine infrastructure contexts. This study investigates the causes and project performance consequences of variation orders in Omani fishery port construction projects using Partial Least Squares Structural Equation Modelling (PLS-SEM). Seven theoretically grounded VO cause categories, contractor-related (CON), client-related (CR), consultant-related (CS), environmental-related (ER), innovation-related (IR), organizational-related (OR), and resource-related (RR), were modeled as independent predictors of project performance (PP), operationalised across cost, time, quality, and stakeholder satisfaction dimensions. Survey data were collected from 113 construction professionals with direct fishery port project experience in Oman, encompassing government client agencies, engineering consultancies, and contracting organisations. The structural model results reveal that contractor-related causes (? = 0.549, p = 0.006, f² = 0.547) and client-related causes (? = 0.281, p = 0.021, f² = 0.280) are the only statistically significant direct predictors of project performance, collectively demonstrating that VO impacts are concentrated in execution capability and client governance rather than distributed evenly across causal domains. The remaining five constructs did not attain direct significance, suggesting their performance effects operate through mediated pathways via contractor execution and client decision-making mechanisms. These findings contribute an integrated, context-sensitive structural model of VO impact to the construction management literature and offer targeted governance recommendations for procuring authorities, engineering consultants, and contracting firms operating within Oman's marine infrastructure program.

Abstract

The study comprehensively investigates hydrodynamic performance, parameter optimization, and scaling implications of a 10-inch-high Cone-Enhanced Split Reaction Turbine (CESRT). By employing a rigorously structured Central Composite Design (CCD), the analysis evaluates the independent, quadratic, and interactive effects of three critical operational parameters: applied torque, system bypass angle, and internal cone size. The statistical model demonstrated exceptional predictive reliability, characterized by an F-value of 319.59 and a coefficient of determination (R²) of 0.9976. The empirical findings reveal that while applied torque serves as the primary driver of operational efficiency, complex, non-linear interactions exist between the torque, the bypass angle governing the inlet flow, and the geometric influence of the internal cone modifier. Maximum hydraulic efficiency for the 10-inch architecture is achieved at an applied torque of 60 N-m, a bypass angle of 22.5°, and a cone size of 2.5 inches, yielding a peak predicted efficiency of 64.5803%. Crucially, a comparative scaling analysis against preceding 8-inch CESRT configurations reveals a significant degradation in isentropic efficiency associated with the increased aspect ratio of the 10-inch model. These findings demonstrate mathematically that radial expansion, rather than vertical elongation, is hydrodynamically optimal for simple reaction turbomachinery.

Abstract

In this paper, a parameter and corresponding lifetime parameters of the New X-Lindley distribution (NX-LD) are estimated. Collecting complete lifetime data from a population using highly reliable items is costly and time-consuming. An adaptive Type-II progressively hybrid censoring scheme under a constant-stress accelerated life test (ALT) is used to balance the optimal test time and the number of failed items. Under the constant-stress ALT model, a variety of statistical methods are developed to estimate the parameters of the lifetime distribution. These methods include maximum-likelihood and Bayesian point estimates. Asymptotic maximum-likelihood, bootstrap confidence intervals, and equal two-sided credible intervals are also formulated. These methods aim to obtain efficient estimators for the distribution parameter, acceleration factor, and reliability and failure rate indices. The estimators are assessed in terms of mean squared error, coverage percentage, and interval length through a Monte Carlo simulation study. The usefulness of the accelerated adaptive censoring scheme is demonstrated with a real-life application in the field of engineering, the voltage and failure time data, as well as a simulated data example to illustrate its practicality.

Abstract

Repurchase intention is a common outcome variable used to link marketing activities to customer equity and firm performance, but bibliometric evidence on this variable is still fragmented over context-specific streams (e.g., service recovery, social commerce, green marketing, and e-commerce). This paper presents a domain-level bibliometric analysis of the literature on repurchase and repeat purchase intention studies across markets and theories. Using Scopus as the sole data source, we searched for publications between 2015 and 2024, and used screening to filter for marketing and consumer behavior studies, yielding a final set of 1,025 documents. We integrate performance analysis with science mapping methods, using reference co-citation, document bibliographic coupling, and co-word analysis, implemented in VOSviewer and facilitated by Bibliometric. Findings indicate a robust trajectory of growth over the past decade with a production pattern that is predominantly Asia-Pacific in context, and a visibility benefit in Anglo-Saxon citation patterns. The maps indicate robust intellectual foundations centered on satisfaction-loyalty process theories, service quality and recovery theories, digital relationship quality, and brand theories, while pointing to new foundations in social and live commerce, AI-driven services, and green consumption. The bibliometric analysis concludes with an integrated research agenda for theory development, methodological innovation, and boundary condition testing.

Abstract

This study addresses the challenge of sustaining employee performance in digital start-ups, where rapid technological change, uncertainty, and highly dynamic work environments often affect employees’ productivity, engagement, and well-being over time. Despite the growing importance of sustainable employee performance, limited attention has been given to the mechanisms through which leadership influences long-term employee outcomes in digital start-ups. In particular, the role of employee dynamic capability as an adaptive mechanism linking change leadership and sustainable employee performance remains underexplored. Therefore, this study examines the effect of change leadership on sustainable employee performance in digital start-ups, with employee dynamic capability as a mediating variable. This study employs a quantitative research design, where data were collected from 329 employees working in digital start-ups and analyzed using Partial Least Squares Structural Equation Modelling (PLS-SEM). The results reveal that change leadership has a strong and significant positive effect on sustainable employee performance. The mediation analysis indicates that employee dynamic capability partially mediates the relationship between change leadership and sustainable employee performance. These findings highlight the importance of leadership practices that not only drive immediate performance outcomes but also strengthen employees’ long-term adaptive capabilities. The study contributes to the literature by integrating leadership and dynamic capability perspectives in explaining sustainable employee performance within the context of digital start-ups. Practical implications suggest that organizations should invest in change-oriented leadership development and foster environments that support continuous learning and flexibility.

Abstract

In light of technological progress and the significant development in the field of artificial intelligence, it has become necessary to apply these technologies in various practical fields, especially in the educational sector in Middle Eastern universities, which still face constraints in technological infrastructure. To address this problem, deep learning–based AI algorithms were used to perform face detection and recognition. These algorithms have been applied to detect and record attendance, and provide daily, weekly and monthly reports. The system uses an application and comparison of the results of applying the MTCNN algorithm for face detection with the FaceNet algorithm for face recognition. The other aspect represents the application of the RetinaFace algorithm with the ArcFace algorithm, which is characterized by its accuracy and efficiency in dealing with difficult circumstances. Finally it was merged

RetinaFace for face detection and MTCNN for facial feature alignment and enhancement, followed by ArcFace for accurate facial recognition. The hybrid combination of these three algorithms proved the accuracy and efficiency of the second algorithm, RetinaFace, with the ArcFace algorithm, because the system gave the same results when applied alone or when combined. Thus, when comparing the results of the three operations, the RetinaFace algorithm with ArcFace proved its superiority over the rest of the proposed algorithms. The system was also designed to work under different environmental conditions and in educational institutions with limited technological infrastructure.

Abstract

This research examines the maturity level of gamification in promoting joyful and mindful learning among 418 senior high school students in Palu, Central Sulawesi, utilising the G.E.A.R. framework and its connection to SDG 4 (Quality Education). Utilising a descriptive-comparative survey strategy, the study indicates that all schools are situated at Level 3.0 (Engaged Integrators), achieving only 64–67% of the optimal score. Urban schools had somewhat higher perception means than rural schools (M = 76.21 vs. 74.03), and the differences between the two types of schools were statistically significant (F(4,413) = 7.14, p < .001). Collaboration and reflection were the strongest aspects, while competition and autonomy were the worst, especially in rural areas. The results show that there is still a plateau at middle maturity and differences between urban and rural areas, but they also show that there are pockets of resilience in rural schools. This study presents a localised urban-rural gamification model for developing contexts and advocates for specific “breakout strategies” emphasising autonomy, equitable competition, and infrastructure fairness to advance practices to Level 4.0 and expedite the attainment of inclusive Sustainable Development Goal 4.

Abstract

Breast cancer classification plays a critical role in early diagnosis and clinical decision-making. However, medical datasets are often high-dimensional and contaminated with outliers, which can degrade classification performance. While Principal Component Analysis (PCA) is commonly used for dimensionality reduction, its sensitivity to outliers limits its effectiveness in medical data analysis. To address this limitation, this study proposes a robust PCA (RPCA) approach based on the Minimum Regularized Determinant (MRCD) estimator for dimensionality reduction and named the proposed method as Robust MRCD-PCA (RMPCA). The classification using proposed RMPCA is evaluated using Support Vector Machine (SVM), Artificial Neural Network (ANN), and K-means classifiers, and compared against baseline and PCA-based models. A total of nine classification models is examined using a breast cancer dataset, with performance assessed using accuracy, precision, recall, and F1-score. Experimental results demonstrate that RMPCA achieves more consistent and reliable classification performance, particularly when combined with supervised classifiers such as SVM, outperforming both baseline and PCA-based approaches. These findings highlight the importance of robust dimensionality reduction as an effective preprocessing strategy for improving machine learning-based breast cancer classification.

Abstract

This study aimed to develop and validate the Malaysian Geography Teachers’ Professionalism Questionnaire Instrument based on the 2027 School Curriculum Framework. The instrument was designed to measure four key dimensions of geography teachers’ professionalism: pedagogical practice, content mastery, assessment implementation, and character education. A quantitative survey design was employed involving 300 secondary school geography teachers in Malaysia. Respondents were selected using proportional stratified sampling and simple random sampling. The questionnaire consisted of 60 items, with 15 items representing each construct. Content validity was established through expert review, while construct validity was examined using exploratory factor analysis. The results indicated that the data were suitable for factor analysis, with a Kaiser-Meyer-Olkin value of .937 and a significant Bartlett’s test of sphericity. Four factors were extracted, explaining 49.161% of the total variance. Most items recorded acceptable factor loadings ranging from .342 to .807, supporting the construct validity of the instrument. Overall, the study produced a valid instrument for measuring geography teachers’ professionalism in the Malaysian secondary school context and provides a useful tool for researchers, policymakers, and educational stakeholders.

Abstract

Digital crowdsourcing platforms depend on freelancer profiles to support trust, reputation, and hiring decisions. However, anomalous profiles with unusual attribute combinations may reduce platform reliability. Existing anomaly detection studies mainly emphasize detection accuracy while overlooking robustness, interpretability, and governance relevance. This study proposes an explainable anomaly detection framework for identifying anomalous freelancer profiles using Isolation Forest (IF) and Local Outlier Factor (LOF). Hyperparameter optimization techniques, including Grid Search, Randomized Search, Genetic Algorithms, and Bayesian Optimization, are applied to improve model robustness. Beyond detection performance, the framework evaluates anomaly consistency, clustering separability, and cross-model agreement. To improve transparency, SHAP-based explainability is integrated to interpret anomaly scores and identify feature-level contributions. This enables platform administrators to understand why freelancer profiles are classified as anomalous and supports human-in-the-loop decision-making. Experimental results show that optimized models consistently detect economically and reputationally implausible profiles while reducing unstable detections. The findings demonstrate that integrating optimization and explainable AI enhances trustworthy anomaly detection and supports governance in digital crowdsourcing platforms.

Abstract

The high number of applicants for a job vacancy makes it difficult for the Human Resources (HR) team to find the best candidates who can be called for the interview process. This highlights the need for computational analysis that can semantically understand and compare applicant responses. This research implements Natural Language Processing into an artificial intelligence (AI)-based model that is capable of selecting applicant candidates in open-ended questions. Data from 100 applicants served as input. The NLP model was empirically tested by reviewing the recommendations provided and the candidates' responses. The model proved effective in assisting HR in candidate selection.

Abstract

The rapid advancement of Generative Artificial Intelligence (GenAI) has significantly influenced educational assessment practices, including the teaching and assessment of the Malay Language. The integration of this technology needs a strong ethical understanding among teachers to preserve the integrity, fairness, and authenticity of assessment processes.In line with this, this study aims to evaluate the content validity of an instrument designed to measure teachers’ understanding of GenAI ethics in the context of Malay Language assessment through expert panel evaluation. This study employs a quantitative research design involving five expert panellists, consisting of professional experts and a lay expert, who assessed the instrument items in terms of content relevance, construction, and language clarity. The instrument covers six core GenAI ethics constructs: intellectual property, truthfulness, robustness, malicious use recognition, sociocultural responsibility and human-centric design. Content validity was examined using the Content Validity Ratio (CVR) method based on Lawshe’s (1975) guidelines with three likert points scale. The findings reveal that 50 items achieved the maximum CVR value and were accepted by the expert panel, while only 12 number required refinement to improve clarity and construct alignment. Overall, the results provide strong evidence that the instrument demonstrates high content validity and is suitable for assessing Malay Language teachers’ understanding of ethical GenAI use in assessment. However, this study is limited by the relatively small number of expert panellists involved in the validation process and the focus on content validity alone. Future research is recommended to conduct pilot testing with a larger sample of teachers and to examine additional psychometric properties, such as construct validity and reliability, using statistical approaches such as exploratory factor analysis or Rasch measurement analysis.

Abstract

This study aims to analyze the validity of e-modules for mathematics learning in the Ammatoa cultural context, to support students' creative thinking skills in the subject matter of flat-sided cubes and blocks. The research method applied is research and development (R&D) using the Plomp development model, which consists of three phases: the preliminary research, the prototyping phase, and the assessment phase. However, the assessment phase in the form of a trial has not been carried out because this study is limited to the prototype phase, which focuses on testing the validity of the content and media of the developed e-module. The data analysis method involved qualitative and quantitative descriptive analysis, where qualitative descriptive analysis was used to process reviews from content and media experts to revise the developed product, while quantitative analysis was used to process data obtained through expert validation sheets. The validity of the use of Aiken's V scale was evaluated. The results of this study indicate that the mathematics learning e-module contextualized in Ammatoa culture is categorized as valid, with an average content validity score of 0,731, while the media/design validity score is 0,733, requiring minor revisions, and is deemed suitable for use. This electronic module is presented through a flipbook platform, allowing students to access it easily anytime and anywhere, either independently or in groups, which supports the development of students' creative thinking skills.

Abstract

Financial time series forecasting faces challenges from structural instability caused by macroeconomic shocks. The hybrid Autoregressive Integrated Moving Average-Generalized Autoregressive Conditional heteroskedasticity models (ARIMA-GARCH models) can capture temporal dependencies and volatility in financial time series but tend to produce inaccurate forecasts as they assume constant model parameters across structural breaks in the series. This study develops a forecasting framework that incorporates structural break parameters into the ARIMA-GARCH model. The study uses monthly closing price of S&P 500 index from January 2001 to December 2017. The Bai-Perron test is used to identify the structural breaks in the time series, followed by modelling the ARIMA mean and GARCH variance of the series in each regime. The structural break in early 2009 is identified as a transition from a moving-average process in the pre-crisis period to a random walk process in the post-crisis period. Simulations studies indicate that the proposed framework significantly reduces distortions. The Mean Absolute Percentage Error (MAPE) in the forecast reduces from 59.8% in the full-sample ARIMA-GARCH model to 13.0% in the regime-adaptive model. The regime-adaptive model developed in this study offers a better tool for forecasting the S&P 500 index, particularly during periods of high volatility.

Abstract

Background: In many of southern Iraq's rural and semi-arid areas, groundwater serves as the primary supply of drinking and household water. However, inappropriate waste disposal, agricultural practices, and natural geochemical processes—which can result in microbial and chemical pollution—are rapidly endangering its purity.  The study's objective is to provide a thorough assessment of the water safety for human consumption by evaluating the physicochemical and biological quality of groundwater in the Al-Salman district of Iraq by identifying bacterial contaminants, analyzing their antibiotic resistance, and measuring important chemical parameters. Material and methods In the spring of 2024, groundwater samples were taken from two sites in the Al-Salman district using sterile bottles (100 mL for bacteriological examination and 1 L for chemical analysis). Standard analytical techniques were used to determine physicochemical parameters including pH, temperature, electrical conductivity (EC), total dissolved solids (TDS), alkalinity, nitrate, nitrite, sodium, potassium, and phosphorus. Samples were enriched in Brain Heart Infusion (BHI) broth and incubated for 24 hours at 37 °C before being subculture on MacConkey agar, blood agar, and Mannitol salt agar for bacteriological examination. Biochemical assays and colony morphology were used to identify the bacterial isolates. The Kirby-Bauer disk diffusion method was used to test for antibiotic susceptibility against a variety of antibiotics in accordance with CLSI standards. Results: According to the physicochemical investigation, the groundwater had a nearly neutral pH ranging from 7.00 to 7.10 and relatively high temperatures between 30 and 33?°C, which may promote microbial growth. Electrical conductivity (EC) and total dissolved solids (TDS) were very high, indicating a large amount of salt and minerals in the water. Alkalinity (ALK) was within permissible limits at less than 250?mg/L. Nitrate (NO??) levels exceeded the allowable limit of 50?mg/L, while nitrite (NO??) remained within the safe range around 50?mg/L. Potassium (K?) concentrations were higher than the recommended limit of 10–15?mg/L, and sodium (Na?) was also elevated above 250?mg/L, whereas phosphorus (P) was within acceptable levels around 0.4?mg/. The biological analysis showed several pathogenic bacteria were isolated, including Escherichia coli, Staphylococcus aureus, Klebsiella pneumoniae, Staphylococcus epidermidis, Proteus vulgaris, Enterobacter spp., and Pseudomonas aeruginosa, many of which exhibited multidrug resistance. However, antibiotics such as meropenem, gentamicin, amikacin, and netilmicin showed the highest effectiveness against most isolates. Conclusion: These findings indicate that groundwater in the studied area is unsuitable for direct drinking without proper treatment and continuous monitoring to minimize potential public health risks.

Abstract

The rapid development of educational technology offers new opportunities to enhance creative writing instruction, including drama script writing, which remains challenging due to limited instructional scaffolding and culturally relevant resources. This study aimed to develop and evaluate FlipHTML5-based digital learning media for drama script writing grounded in South Sumatra folktales to promote creative and culturally responsive learning in higher education. This research employed a Research and Development (R&D) approach adapted from Alessi and Trollip, covering the planning, design, and development stages. Data were collected through needs analysis questionnaires, expert validation (alpha testing), and user evaluation (beta testing). Product feasibility was assessed by content, media, and language experts, while effectiveness was measured using a one-group pretest–posttest design involving students enrolled in a drama writing course. The results show that the media are highly feasible in terms of content (87.5%) and language (91.6%), and feasible in media design (75%). Student responses indicate high usability, engagement, and pedagogical relevance. Statistical analysis revealed significant improvement in students’ drama writing skills, with mean scores increasing from 53.03 (pre-test) to 85.42 (post-test) (p < 0.05). These findings suggest that integrating dramaturgical theory, interactive digital features, and local folklore effectively enhances creative writing skills while supporting cultural preservation. The developed media function not only as a digital resource but also as a platform for creative production in literature instruction.

Abstract

Activities related to the quality assurance (QA) have become an indispensable element in the improvement of higher education (HE) effectiveness and competitiveness. Although significant investments have been made in quality assurance systems it is still under study how much these practices can contribute to an institutions performance. Hence, the focus of this study is on how the quality assurance (QA) practices affect institutional performance in higher education institutions, focusing on the strategic leadership and organizational culture dimensions. A Quantitative approach that includes cross sectional survey design was used. The subjects were faculty members, academic administration and Quality Enhancement Cell (QEC) staff of public sector Universities of Saudi Arabia. Stratified random sampling method was used to select a sample of 400 for the study. A structured questionnaires designed and based on validated instruments were used to collect data on the practice of quality assurance, strategic leadership, organizational culture and institutional performance. Structural Equation Modeling (SEM) using Smart PLS was used as an analysis method for the data collected. The results showed that the quality assurance practices had a significant positive effect on institutional performance. The findings also suggested that strategic leadership has a huge role in strengthening the relationship between quality assurance practices and the institutions' performances by enabling the effective implementation of the quality initiatives. Furthermore, it was discovered that the practices of quality assurance are significantly affected by organizational culture, thus creating an environment conducive to continuous improvement and institutional excellence. The scope of the study is an addition to existing literature on higher education management, offering empirical insights on the significance of quality dealing with, strategic management and organizational culture for better higher education management. The results hold implications for settings such as university administrations, university policymakers and quality assurance practitioners wishing to improve institutional effectiveness, sustainability, academic excellence, and enabling strategic quality management practices.

Abstract

Subjective well-being (SWB) is individuals’ cognitive and emotional evaluations of their lives, including life satisfaction and affective experiences. However, limited research has examined the relationship between meaning in life and SWB and the role of hope in this relationship among persons with visual impairment, particularly in Malaysia.  This study investigates the role of hope as a mediator in the linkages between meaning in life and subjective well-being (SWB) among visually impaired individuals in Malaysia. Using a cross-sectional design, data were collected from 144 participants, member of the Society of Blind Malaysia (SBM) and the Malaysian Association for the Blind (MAB). Meaning in life was assessed using the Meaning in Life Questionnaire (MLQ) and hope was measured using the Adult Hope Scale (AHS) respectively, while SWB was assessed using a multidimensional instrument adapted from established quality of life and well-being measures. Correlational analysis showed significant positive relationship between meaning in life, hope and SWB. The PROCESS Macro mediation analysis indicated that hope partially mediates the relationship between meaning in life and SWB. Specifically, those who felt a greater sense of life meaning showed more hope, which was linked to enhanced SWB. These results highlight the significance of cultivating hope in psychological treatments to improve the well-being of individuals with visual impairments. The study provides actionable insights for social support services and emphasizes paths for upcoming studies in positive psychology and disability research.

Abstract

The purpose of this study is to describe the development of a mechanical wheeled robot car (using Mecanum wheels) that can follow an object via Computer Vision (CV). The object tracked is a regular 2-dimensional shape with a specific color (yellow), which can be attached to a person or moving object and followed by the robot at a set distance. The research methodology uses thr  CV process, which begins by identifying the color of the object in the HSV (Hue, Saturation, and Value) color space, followed by shape recognition based on the number of contour keypoints obtained through Structural Analysis and Shape Descriptors. The outcomes of this combination enable the recognition of objects with a specific shape and color. To maintain the distance between the object and the robot, a calibration procedure is performed, and the distance between the robot's camera and the object is calculated using mathematical 2D distance calculations. This robot was built with a Single Board Computer (SBC) Raspberry Pi 4B+ controller (4GB memory and 32GB external memory), a Pi camera as the vision sensor, a Debian 11.0 "Bullseye" operating system, and a Python application, with assistance from the OpenCV library. According to test results on a yellow triangle-shaped object, the robot is able to track the movement of the object effectively, as intended, despite interference from objects with triangular shapes of different colors or objects of the same color.