In recent years, significant developments have occurred in AI-based chatbots that have been effectively deployed in the educational field. However, given the novelty of this technology, descriptive analyses remain scarce. Although many review studies have focused on the effectiveness of chatbots, they generally present broad results, and only a few have addressed the impact of this technology on learning outcomes. The present study examines the educational implications of AI chatbots on various learning outcomes through a post hoc analysis conducted in accordance with PRISMA principles. It aims to aggregate and analyze findings from studies that examined the use of chatbots and their impact on specific learning outcomes. A total of 26 studies were selected from a pool of 6,721 published between 2021 and 2024 and indexed in the Scopus and Web of Science databases. Data analysis was conducted using the Newcastle-Ottawa Scale (NOS) for Education. The results revealed that AI-chatbot technology has a positive influence on several learning outcomes, including academic achievement, motivation, self-assessment, engagement in learning, self-efficacy, and language learning, among others. The studies also detailed the methodologies and tools employed in these investigations. The study also offers insights into how intelligent chatbots can be leveraged to enhance various learning outcomes.

Vocational education plays a pivotal role in nurturing talent and supporting national development. However, challenges such as outdated talent development concepts, insufficient teacher training, and a lack of attention to soft skills cultivation from both schools and students have hindered the comprehensive development of secondary vocational students. This study aims to explore the direct effect of perceived teachers’ transformational leadership on the soft skills of 324 secondary vocational students in China and to examine the mediating role of students’ self-efficacy in this relationship. Using Structural Equation Modeling (SEM), the results show that teachers’ transformational leadership has a significant positive effect on students’ soft skills (β = 0.33, p < .01). Moreover, self-efficacy partially mediates this relationship (indirect effect β = 0.07, p < .05), accounting for 22.6% of the total effect. These findings suggest that teachers’ inspirational motivation, individualized consideration, and intellectual stimulation directly foster students’ communication, teamwork, and problem-solving skills, while also indirectly strengthening them by enhancing students’ confidence. Practically, the study underscores the need for teacher training in transformational leadership and for policy initiatives that integrate soft skills into vocational curricula.

This study aimed to examine the impact of self-organization and self-efficacy in predicting attitudes toward cheating in online exams among undergraduate students. Drawing on the self-regulated learning framework, a cross-sectional survey design was employed to collect data from 153 students at a Saudi university using three validated scales. The findings revealed that self-organization and self-efficacy were significant negative predictors of attitudes toward cheating, with higher levels of these attributes associated with less favorable attitudes toward cheating. Furthermore, moderation analysis showed that self-efficacy significantly moderated the relationship between self-organization and cheating attitudes, suggesting a complex interplay between the two constructs. The findings highlight the importance of fostering self-regulated learning skills to enhance academic integrity in online assessment contexts.  

The introduction of evolution in post-apartheid South Africa in 2008 presented significant challenges for teachers, due to both their own and their students' faith-based biases. This study investigates South African teachers' attitudes and understanding of evolution, utilizing Lemkian sociocultural theory and Vygotskian ZPD scaffolding. A mixed-methods approach was employed, including a Likert-scale questionnaire from 91 life science teachers and qualitative insights from open-ended questions. Results reveal a lack of enthusiasm for teaching evolution, influenced by personal religious beliefs and doubts about its scientific validity. Teachers often call for equal representation of evolution and creationism in classrooms. The preparation of teachers must address the stagnant discussion of evolution, while the life sciences curriculum emphasizes the Nature of Science, scientific literacy, and connections between science and society is compromised.

This study aims to address students' low mathematical problem-solving skills and self-regulated learning by developing website-based mathematics learning media using the problem-based learning model. Learning that integrates website-based media allows for more independent learning activities in solving mathematical problems. This study uses the ADDIE model to test the validity, practicality, and effectiveness of the product. Website-based media met the validity criteria based on expert assessments with an average score of 137 from subject matter experts and 78.5 from media experts, and the research instruments were declared valid with an Aiken index ranging from 0.9 to 0.99. Website-based media meet the criteria for practicality based on an average teacher assessment of 90.5, an average student assessment of 50.1, and a learning implementation rate of 93%. Website-based media are proven to be effective through four indicators: (1) students who pass exceed the threshold of 80%. (2) One sample t-test results p-value <.05 and t-value >1.69, so that H_0 is rejected. This proves that website-based mathematics learning media are effective in improving students' mathematical problem-solving skills and self-regulated learning. (3) The results of the paired sample t-test show p-value <.05 and t-value >1.69, so that H_0 is rejected. This proves that there is a significant improvement in students' mathematical problem-solving and self-regulated learning abilities after using website-based mathematical learning media. This study is able to facilitate students' problem-solving and self-regulated learning abilities simultaneously in mathematics education.

Team-based learning (TBL) is a popular form of collaborative learning designed to increase student engagement and motivate students to learn. A growing body of research, particularly in the health sciences, has demonstrated that TBL has positive impacts on student performance and classroom dynamics. However, much less is known about the student outcomes associated with TBL courses in undergraduate science, technology, engineering, and mathematics (STEM) education, fields in which active learning is especially important for student success and retention. To address this gap, we conducted a systematic review of the student outcomes associated with TBL in undergraduate STEM education. Using the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) method, we identified 55 empirical and qualitative research articles published between 2005 and 2024 that reported on TBL implementation practices and student outcomes. Importantly, we found that most studies described increased student performance and improved measures of classroom dynamics when TBL was compared to lecture-based teaching approaches. These findings provide further evidence that TBL is an effective instructional method and suggest that TBL can be implemented successfully across a wide range of student populations and undergraduate STEM disciplines.

Teacher-targeted victimization remains an under-integrated dimension of school violence research. This systematic review follows PRISMA 2020 guidelines and synthesizes 56 empirical studies (2015–2026) examining how teacher-targeted victimization is conceptualized and analyzed. The review maps patterns in geographic distribution, methodology, behavioral aggression forms, perpetrator framing, and the integration of institutional variables. Findings reveal the predominance of quantitative cross-sectional designs (75.9%). Physical (n = 28) and verbal aggression (n = 26) are the most frequently examined forms, with over half of the studies conceptualizing aggression as exclusively student-perpetrated. Institutional and governance variables remain limited: 84.8% of studies include no institutional variables beyond exposure or rely solely on general school climate indicators, while governance-level constructs appear in only one study. Drawing on Organizational Justice Theory and Institutional Theory, the review advances a Multilevel Institutional Accountability Model that conceptualizes teacher-directed aggression across behavioral exposure, institutional processing, and governance architecture. The findings highlight the need for greater integration of institutional and governance variables in future research.