Implementing the experiential learning model is one effective step in improving student learning outcomes, but recent evidence regarding its effectiveness is still inconsistent and not widely reported. Therefore, to gain a deep understanding and identify the effectiveness of experiential learning practices, this meta-analysis review was conducted. This review was conducted on 23 studies with 40 effect sizes obtained from the Scopus, ERIC, and Web of Science databases. The analysis model used was a random effects model with a robust variance estimation (RVE) approach to address the issue of interdependence between effect sizes within a single study. The measurement results show a high overall effect size (Hedges's g = 1.15; 95% CI), indicating a significant positive impact. Heterogeneity estimates are very high (I² = 98.2%), requiring further moderator analysis. Meta-regression analysis of the variables of education level, treatment type, and learning outcome dimension showed that only the education level variable approached significance in moderating the variation in learning outcomes. Subgroup analysis showed variation in the effectiveness of intervention implementation within each moderator category, but all were within the positive range. Publication bias tests showed bias, but correction using the trim-and-fill method and Peter's test proved that these findings remained consistent and valid. Overall, experiential learning has proven effective in enhancing and shaping various dimensions of student learning outcomes. This approach has great potential to continue being used and integrated into future learning practices, especially through the support of various pedagogical innovations and technology integration.

Traditional chemical engineering education faces persistent challenges, as lecture-based instruction and limited laboratory access restrict students’ engagement, practical skill development, and comprehension of complex engineering concepts. In overcoming these challenges, immersive virtual reality (VR) learning environments are increasingly being adopted to enhance students’ visualization, interactivity, and experiential learning. Therefore, this study aims to explore how VR can transform chemical engineering education by enhancing student engagement, conceptual understanding, and practical learning experiences. This review also analyzed and mapped research trends in the application of VR for immersive learning in chemical engineering education using keywords like ''Immersive Learning'', ''Virtual Reality'', ''Chemical Engineering'' and ''Chemical Engineering Education''. Accordingly, the keyword co-occurrence analysis revealed four thematic clusters linked to immersive visualisation, collaborative VR learning, VR–AI integration, and safety-oriented training. Based on these findings, a four-stage curriculum integration model is proposed (pre-conceptual familiarisation - immersive experimentation - hybrid transfer - real-lab validation). A comparative cost analysis indicates that although VR-based learning demands a higher upfront investment, it achieves cost parity within approximately 1.5 years and reduces total training expenses by about 40–45% in the third year, offering greater economic advantage for larger student cohorts. Finally, the synthesis indicates that VR can enhance conceptual understanding, hazard-awareness, and systems-level reasoning while improving utilisation efficiency in laboratory-intensive programmes. Future research should prioritise controlled cohort comparisons and longitudinal verification of transferability to physical plant behaviour.

Project-Based Learning (PjBL) is often proposed as a way to engage students through authentic tasks, yet classroom observations in junior high schools in Malang indicated that students’ critical thinking and learning motivation were still relatively low. This study examined whether PjBL supported by a mobile application could improve seventh-grade students’ critical thinking and learning motivation on the topic of substances and their changes. A quasi-experimental design was used with 62 students from one junior high school (30 in the experimental group and 32 in the control group). The intervention combined PjBL with a mobile application and supporting materials (modules and student worksheets). Expert review showed that the learning media met acceptable validity criteria, with an average score above 83 percent (high validity). The instruments demonstrated good internal consistency, with Cronbach’s alpha of 0.853 for the critical thinking test and 0.929 for the learning motivation questionnaire. Independent-samples t-tests indicated statistically significant differences between groups (p < 0.05). The experimental group scored higher than the control group in critical thinking (t(60) = 2.33, p = 0.023, d = 1.02) and learning motivation (t(60) = 11.83, p < 0.001, d = 1.13). These results suggest that, in this context, mobile-aided PjBL was associated with higher critical thinking and learning motivation than conventional instruction, and it may be considered for teaching substances and their changes.

Social concern represents the moral and behavioral manifestation of empathy, responsibility, and prosocial engagement in addressing societal problems. Religiosity, as a deeply internalized system of belief and moral commitment, has long been recognized as a potential source of prosocial motivation. Meanwhile, social media has emerged as both a powerful channel for mobilizing collective compassion and a potential driver of self-centered or performative behavior. This study examines how religiosity moderates the relationship between social media use and social concern among senior high school students in Malang City, Indonesia. Employing a quantitative explanatory design with a sample of 414 respondents selected through stratified sampling, data were collected using a structured questionnaire and analyzed through Moderated Regression Analysis (MRA). After reanalysis and correction, the regression model shows that religiosity significantly strengthens the positive effect of social media use on social concern. The adjusted coefficient of determination (R²) increased modestly from 0.04 to 0.07 after including the interaction term, indicating that religiosity serves as a meaningful moderator in this relationship. This finding suggests that adolescents with higher religiosity levels are more likely to transform social media exposure into altruistic and socially responsible behavior. The study contributes to the sociology of religion and communication studies by offering an empirical model linking digital engagement and faith-based values in promoting social solidarity.

Based on Expectancy-Value Theory, talent and interest are often assumed to relate to individual competence. However, few studies have examined talent, interest, and competence simultaneously using SEM-PLS. This study analyzed the direct and indirect relationships among talent, interest, and the teaching competence of prospective mathematics teachers using SEM-PLS. The participants were 87 fifth-semester students in the Mathematics Education Department at UIN Siber Syekh Nurjati Cirebon. Data were collected through an interest questionnaire, a talent test, and teaching-competence observations. The results indicated that talent had a positive but non-significant effect on interest (coefficient = 0.487; p = 0.127) and on teaching competence (coefficient = 0.236; p = 0.531). Interest showed a negative but non-significant effect on teaching competence (coefficient = -0.243; p = 0.289). The mediation test also showed that interest did not significantly mediate the effect of talent on teaching competence (indirect coefficient = -0.118; p = 0.369). The explained variance was 0.237 for interest and 0.059 for competence, indicating that the model accounted for only a small proportion of variance. These findings suggest that teaching competence is more likely shaped through experience and social interaction in educational settings than as a direct outcome of talent or interest. The lack of significant relationships highlights the complexity of teaching-competence development. Practically, teacher education institutions should support talent and interest exploration while strengthening competence development through structured mentoring, early career exploration workshops, and systematic talent identification.

This research develops a virtual reality-based game with an ethnomathematics approach to Batam City culture in the form of Luti Gendang food in learning social arithmetic material in mathematics learning at the elementary school level. This product's development is significant because students are now a digitally native generation and are more interested in learning through technological media than traditional media. Using the ADDIE model as a basis, this research will focus on the development stage, which includes 3 steps, namely (1) initial product design in the form of making storyboards, making learning videos with Adobe Premiere Pro, designing image designs for VR displays, and making learning videos; (2) initial product development focuses on making VR games and conducting validation tests by experts covering the feasibility of material, media, and language to determine the feasibility of the media being developed, and (3) refining the initial product by making revisions and improvements based on the results of the validation test. The development findings demonstrate that: (1) The initial product design incorporates five key menus—VR Simulation, Material, Video, Quiz, and Info; (2) The expert evaluations of the product’s feasibility across media, language, and content aspects yielded average scores of 85.2, 80.8, and 83.1, respectively, classifying the product’s overall feasibility as excellent; and (3) Subsequent refinements were implemented based on expert validation and user testing feedback, specifically targeting interactivity enhancements, improvements to VR visual elements, and adjustments to in-VR narrations and directional cues.

Ecosystem learning in secondary education often relies on theoretical instruction with minimal contextual engagement, leading to persistent misconceptions and low student understanding. This study aims to address this pedagogical gap by developing the Riverkeeper game, an innovative hybrid learning media that combines digital platforms (Canva, Wordwall, and Google Docs) with field exploration activities to support ecosystem instruction. Using a modified 4D development model (Define, Design, Develop, Disseminate) but only Define, Design and Develop stages were conducted in this paper. Data were collected from five biology teachers and twenty Grade 11 students to identify instructional challenges and needs. The Riverkeeper game was designed based on constructivist and experiential learning principles, integrating digital problem-solving tasks with real-world observation of river ecosystems. The media prototype underwent expert validation involving three evaluators (a biology education lecturer, a media expert, and a senior high school biology teacher) using a structured instrument covering three aspects: game guidance, visual communication and design, and functionality. Validation results showed an overall score of 88% (very valid), with strengths in instructional clarity and visual design, though minor revisions were suggested for navigation and technical integration. A practicality test involving twenty students reported high satisfaction and relevance to learning objectives. As this study focused solely on the development and validation stages, further research is recommended to evaluate the game’s effectiveness in improving student learning outcomes. The Riverkeeper game presents a novel approach to ecosystem education by integrating digital game-based learning with direct environmental exploration.

This study aims to understand the perception and utilization of artificial intelligence (AI) by Catholic Religious Education teachers in learning, as well as identify factors that influence the effectiveness of using this technology. The role of AI in the era of technology 5.0 is not only as a tool, but also as a breakthrough that helps in designing innovative learning. However, many teachers still lack the ability to utilize AI appropriately. Based on this condition, it is necessary to conduct an in-depth study related to teachers' perceptions and their ability to utilize AI in learning, especially in Catholic religious education. A qualitative research method with a phenomenological approach was used to explore the experiences and views of teachers from various levels of tenure. Data were collected through semi-structured interviews, participatory observation, and documentation, then analyzed using thematic analysis techniques and data triangulation to increase the validity of the findings. The results show that there are variations in perceptions related to AI, where teachers with less than five years of service tend to be more open and actively use AI in learning, while teachers with more than 15 years of service show skepticism and limited use. The main influencing factors were technical skills and anxiety related to changing teacher roles as well as the unsuitability of AI in conveying the spiritual and moral values of Catholicism. Training limitations and lack of in-depth understanding of AI were also significant barriers to its utilization. This study highlights the importance of continuous training that integrates technical aspects and religious values to improve teacher readiness. In addition, there is a need to develop AI content that is in line with Catholic teachings as well as learning approaches that emphasize the role of the teacher as a moral and spiritual guide. This research provides a basis for developing teacher professional development strategies and AI-based learning innovations that are aligned with the values of faith in religious education.

This study aimed to develop an assessment instrument to measure students’ ill-structured problem-solving skills using local socio-scientific issues (SSI) on environmental pollution. The research employed a Research and Development (R&D) model with four stages: (1) preliminary study, (2) item development, (3) expert validation, and (4) trial. The instrument was piloted with 120 chemistry education students from UIN Mataram and Universitas Pendidikan Mandalika in Indonesia. The final instrument consisted of 27 essay-based items mapped to four stages of ill-structured problem solving: 8 items for problem representation, 6 for solution generation and selection, 5 for justification, and 7 for solution monitoring and evaluation. Expert validation yielded a high construct validity index (CVI = 0.93). Readability analysis indicated that 75% of students found the language clear, 78% understood the visuals, and 70% were familiar with the chemical terminology. Results indicated that 17 items were of moderate difficulty, 7 were easy, and 3 were difficult. The instrument demonstrated high internal consistency, with a Cronbach’s Alpha value of 0.805. This instrument offers a valid and reliable tool for evaluating students’ engagement in authentic, context-rich chemistry problems. It supports chemistry educators in fostering critical thinking and scientific literacy by embedding local socio-scientific issues into semester-long instruction cycles.

This study explores the effectiveness of diagram construction activities, including area proportional and number line representations, in enhancing conceptual understanding and problem-solving performance in fractions among fifth-grade elementary students. Despite the established benefits of visual models, their systematic application in problem-solving contexts remains underexplored. Many elementary students struggle to integrate symbolic and visual representations in fraction problem-solving meaningfully. Using a quasi-experimental design involving 120 students from four public schools in Makassar, the findings reveal that students who participated in diagrammatic reasoning training significantly outperformed those in the control group in both post-test scores and short-term retention. Specifically, the intervention engaged students in constructing area diagrams and number lines to actively visualize fractions. The integration of spatial visualizations through a dual-coding approach proved effective in strengthening symbolic-spatial connections and reducing extraneous cognitive load, with area-proportional strategies emerging as the strongest predictor of accuracy. These results underscore the importance of routinely integrating diagrammatic activities into elementary mathematics curricula and open new avenues for future research using technologies such as eye-tracking and artificial intelligence to support personalized visual learning.