Computers & Education最新文献

{"title":"Collaborative AI-in-the-loop pedagogical conversational agent to enhance social and cognitive presence in cMOOC","authors":"Jianjun Xiao ,&nbsp;Yulin Tian ,&nbsp;Cixiao Wang","doi":"10.1016/j.compedu.2025.105451","DOIUrl":"10.1016/j.compedu.2025.105451","url":null,"abstract":"<div><div>While connectivist MOOCs (cMOOCs) emphasize learner autonomy and peer interaction, the lack of structured facilitation often hinders the development of social and cognitive presence. This study investigates how Pedagogical Conversational Agents (PCAs) enhance learners' social and cognitive presence in cMOOC environments. Using a within-subjects quasi-experimental design, 344 learners (primarily in-service teachers, 93.60 %) participated in a cMOOC course over four weeks. The 54 discussion topics generated by learners were randomly assigned to experimental (with-PCA, <em>N</em> = 19) and control (without-PCA, <em>N</em> = 35) conditions. A collaborative AI-in-the-loop design was implemented where AI-generated content underwent human review before publication. Analysis of 5301 discussion posts using the Community of Inquiry framework revealed that PCAs significantly enhanced open communication (<em>p &lt; .001, r=-0.213</em>) and group cohesion (<em>p=.027, r=-0.106</em>) in social presence, while improving higher-order cognitive processes, including integration (<em>p=.011, r=-0.119</em>) and resolution (<em>p=.028, r=-0.145</em>). Direct interaction with PCAs yielded superior outcomes in affective expression (<em>p=.027, r=0.103</em>) and specific communication behaviors compared to co-present modes. The findings demonstrate that PCAs provide compensatory support by strengthening cMOOCs' relatively weak components of cognitive presence (integration and resolution) rather than providing comprehensive intervention; direct interaction modes with PCA yield more benefit for deep exploration and social engagement. This offers evidence-based design strategies for AI agent implementation in online education.</div></div>","PeriodicalId":10568,"journal":{"name":"Computers & Education","volume":"240 ","pages":"Article 105451"},"PeriodicalIF":10.5,"publicationDate":"2026-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145047692","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"教育学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"An intelligent recommender system based on K-nearest neighbors to foster self-regulated learning and reduce cognitive load in online higher education","authors":"Shao-Chen Chang ,&nbsp;Ngoc Diep Dao","doi":"10.1016/j.compedu.2025.105470","DOIUrl":"10.1016/j.compedu.2025.105470","url":null,"abstract":"<div><div>Higher education institutions worldwide are increasingly adopting online learning systems to improve instructional quality and offer students flexible, anytime-anywhere access to educational resources. Despite these benefits, online learning environments still face notable challenges. In this learning environment, the major challenge lies in the lack of personalized learning support. The “one-size-fits-all” approach ignores learners' diverse needs, making it difficult for students to regulate their learning effectively. Without tailored guidance, students especially low achievers are more likely to experience cognitive overload, reduced motivation, disengagement, and even dropout. In response to these issues, this study presents an interactive self-regulated learning system built on a comprehensive system architecture that integrates personalized learning path recommendations based on K-nearest Neighbors (KNN) for university students. The system consists of four main components: Content Management System (CMS), Evaluation System, Recommendation System, and User Interface application. This study evaluates the system's perceived effectiveness in the “Introduction to Database Systems” course at a university located in northern Taiwan. It further investigates the system's impact on students' learning behavior, self-regulated learning skills, learning achievement and perceived cognitive load in an online learning environment. The findings suggest that students using the interactive self-regulated learning system with personalized recommendation paths exhibited higher motivation, engagement, learning achievement and self-regulated learning skills. The system also contributed to a noticeable reduction in students'perceived cognitive load.</div></div>","PeriodicalId":10568,"journal":{"name":"Computers & Education","volume":"240 ","pages":"Article 105470"},"PeriodicalIF":10.5,"publicationDate":"2026-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145217982","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"教育学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Identifying effective timer and proctoring settings for online exam systems: Insights from an eye-tracking experiment based on the stimulus-organism-response model","authors":"Tingting Jiang ,&nbsp;Huiyi Tian ,&nbsp;Yanrun Xu ,&nbsp;Han Long","doi":"10.1016/j.compedu.2025.105436","DOIUrl":"10.1016/j.compedu.2025.105436","url":null,"abstract":"<div><div>This study explores how timer and proctoring settings impact examinees’ responses and the underlying mechanism in virtue of the Stimulus-Organism-Response model. A controlled experiment was conducted based on a 2 × 2 within-subjects design, manipulating timer setting (exam timer vs. question timer) and proctor setting (live video feed shown vs. hidden). 49 participants completed mock exams under varying conditions while their attention on the exam questions was captured with an eye tracker. Results revealed exam timers improved user experience across proctoring conditions. The exam timer with showing the live proctoring video feed maximized participants’ total fixation duration on exam questions, whereas the combination of the question timer and hiding the video feed yielded the highest exam scores. State test anxiety was a mediator between online exam system interface settings and exam performance. This work bridges practical and theoretical gaps by offering evidence-based recommendations for designing online exam interfaces.</div></div>","PeriodicalId":10568,"journal":{"name":"Computers & Education","volume":"239 ","pages":"Article 105436"},"PeriodicalIF":10.5,"publicationDate":"2025-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144867363","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"教育学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"A generative AI educational game framework with multi-scaffolding supports workplace competency development","authors":"Pei-Ching Ngu,&nbsp;Chih-Chung Chien,&nbsp;Yen-Ting Ho,&nbsp;Huei-Tse Hou","doi":"10.1016/j.compedu.2025.105421","DOIUrl":"10.1016/j.compedu.2025.105421","url":null,"abstract":"<div><div>This study proposes a game design framework using generative artificial intelligence based non-player character for providing simulated interactions and instant feedback as a metacognitive scaffolding to help learners develop workplace communication skills and stress resistance through situational experience learning. This study investigates learning effectiveness and psychological responses, including flow, perceived fidelity, cognitive load, and qualitative feedback, and specifically analyzes the behavioral patterns of learners interacting with generative artificial intelligence.</div><div>A total of 91 participants were enrolled in this study and divided into three groups: experimental group 1 (generative artificial intelligence interactive metacognitive scaffolding), experimental group 2 (video metacognitive scaffolding), and control group (text metacognitive scaffolding). The results showed that students in the generative artificial intelligence group responded with higher perceived fidelity and were significantly better than the control group with text-based metacognitive scaffolding in terms of learning effectiveness, flow, and germane cognitive load. Behavioral pattern analysis reveals that learners can effectively obtain a lot of positive help in solving tasks through positive interactions with non-player character chatbot.</div><div>This framework and the findings of the study can be used as a reference for related studies in the field of game-based learning in the use of metacognitive scaffolding, contextual learning, and generative AI. Due to the short experimental period, the results may only reflect the short-term learning effects, and the novelty of Gen AI may also cause bias in affecting the learning outcomes, future studies can increase the duration or number of experiences, and explore the effectiveness in other professional domains.</div></div>","PeriodicalId":10568,"journal":{"name":"Computers & Education","volume":"239 ","pages":"Article 105421"},"PeriodicalIF":10.5,"publicationDate":"2025-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144829608","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"教育学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Inoculating students against misleading data visualizations with perceptual training and informative feedback","authors":"Jihyun Rho ,&nbsp;Martina Angela Rau ,&nbsp;Xiaojin Zhu","doi":"10.1016/j.compedu.2025.105413","DOIUrl":"10.1016/j.compedu.2025.105413","url":null,"abstract":"<div><div>Misleading visualizations are a form of misinformation that distort data interpretation, leading to misunderstandings and poor decision-making. Despite their prevalence in both traditional and digital media, there is limited research on effective educational interventions that help students recognize and resist such deceptive visualization techniques. In this study, we developed a perceptual training intervention specifically designed to inoculate students against misleading visualizations. Unlike traditional perceptual trainings that typically involve only correct visuals, trainings for misleading visualizations require the inclusion of erroneous visual examples. While the informative feedback can help students to efficiently recognize such misleading visualizations and extract correct information, it may also disrupt the inductive learning during training, potentially reducing its overall effectiveness. To shed light into these possibilities, we investigated the role of informative feedback in a perceptual training designed for misleading visualizations. We conducted two experiments with large samples of undergraduate students (Experiment 1: N = 252, Experiment 2: N = 244). Experiment 1 tested the short-term effects of different informative feedback types during perceptual training, while Experiment 2 examined both the long-term retention (after one month) and near transfer to novel misleading visualizations. Results showed that informative feedback significantly improved students’ accuracy and efficiency in detecting misleading features within visualizations. Students also transferred these skills to novel misleading visualization types, and critically the benefits of informative feedback were sustained over a month. These results highlight the potential of perceptual trainings with informative feedback to inoculate students against misleading visualizations, offering valuable guidance for the development of educational interventions aimed at mitigating the effects of visual misinformation.</div></div>","PeriodicalId":10568,"journal":{"name":"Computers & Education","volume":"238 ","pages":"Article 105413"},"PeriodicalIF":8.9,"publicationDate":"2025-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144665029","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"教育学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Internet usage inequality among high school students: Patterns, motivational predictors, and educational outcomes","authors":"Xiaofei Li ,&nbsp;Fu Chen ,&nbsp;Chang Lu","doi":"10.1016/j.compedu.2025.105417","DOIUrl":"10.1016/j.compedu.2025.105417","url":null,"abstract":"<div><div>Concerns over Internet usage inequality, particularly excessive usage patterns among adolescents, have intensified due to their potential to widen educational disparities and reinforce existing social inequalities. Many countries have implemented school-wide mobile phone bans to curb inappropriate use during school hours. However, little is known about how such policies influence students' Internet use in out-of-school contexts, or how different cultural and motivational factors help shape these usage patterns. This cross-sectional study examined heterogeneous Internet use patterns within and outside school across school types, under a national mobile phone ban policy, using data from 5160 Chinese high school students (51 % girls, <em>M</em>_age = 15.64, <em>SD</em> = 0.57). Latent profile analyses revealed three usage profiles in elite schools (minimal users, moderate users, and all-time heavy users) and four profiles in non-elite schools, including a distinct group termed “after-school heavy users”. Multinomial logistic regression showed that in non-elite schools, high ICT-related interest and perceived autonomy in ICT use were associated with higher odds of belonging to the all-time heavy users and after-school heavy users, while perceived ICT competence and ICT as a topic in social interaction predicted more balanced profiles. Notably, after-school heavy users exhibited significantly lower academic performance and ICT competencies, identifying them as a potential risk group in the context of digital inequality. These findings extend our understanding of digital inequality by highlighting the importance of usage patterns beyond mere usage time, and offer practical insights for refining ICT policies and interventions aimed at promoting responsible and equitable Internet use.</div></div>","PeriodicalId":10568,"journal":{"name":"Computers & Education","volume":"238 ","pages":"Article 105417"},"PeriodicalIF":10.5,"publicationDate":"2025-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144763978","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"教育学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"From touch to transformation: Value and applications of TUI technology in early childhood development","authors":"Shuang Liang ,&nbsp;Ning Zou","doi":"10.1016/j.compedu.2025.105437","DOIUrl":"10.1016/j.compedu.2025.105437","url":null,"abstract":"<div><div>Tangible User Interface (TUI) technology, by integrating digital information with physical objects, offers a perceptual-based interaction method for children, providing unique advantages in their interaction with computers. This paper reviews the application of TUI technology in early childhood development, focusing particularly on children aged 2–7 years. The research method divides early development into three parts: motor skills, brain and cognitive skills, and social behavior skills. By analyzing the literature, this paper classifies TUI design into three dimensions: embodied interaction, interactive media, and spatial domains. The study finds that TUI design has significant advantages in promoting children's development, especially stimulating initiative, fostering creative thinking, and enhancing communication and collaboration in the real world. The paper also discusses skill development in different types of embodied interaction in TUI design, including sensory traction, movement traction, and social traction. Additionally, it explores the application of various types of interactive media in children's TUIs, such as screen displays, headsets, multimodal sensors, and robots. Finally, the article proposes future research directions, including developing embodied interaction models, multidisciplinary integration and technological innovation, and hybrid-guided design across various domains.</div></div>","PeriodicalId":10568,"journal":{"name":"Computers & Education","volume":"239 ","pages":"Article 105437"},"PeriodicalIF":10.5,"publicationDate":"2025-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144887144","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"教育学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Mobile learning significantly enhances student learning gains: A meta-analysis and research synthesis","authors":"Juan Garzón ,&nbsp;Daniel Burgos ,&nbsp;Kinshuk ,&nbsp;Ahmed Tlili","doi":"10.1016/j.compedu.2025.105415","DOIUrl":"10.1016/j.compedu.2025.105415","url":null,"abstract":"<div><div>Mobile learning has emerged as a strategy with multiple benefits to promote quality education. However, its overall impact on student learning gains and the factors influencing this effect remain unclear. To address this, we conducted a comprehensive meta-analysis of 253 empirical studies evaluating the effects of mobile learning on student learning gains. Additionally, we performed a meta-regression to examine whether substantive features (education field, education level, learning environment, pedagogical approach), methodological characteristics (control treatment, mobile device, sample size), or study quality moderated the effect size. The results showed a large effect of mobile learning on student learning gains <span><math><mrow><mrow><mo>(</mo><mrow><mi>g</mi><mo>=</mo><mn>0.90</mn><mo>,</mo><mi>p</mi><mo>&lt;</mo><mo>.</mo><mn>001</mn></mrow><mo>)</mo></mrow><mtext>.</mtext></mrow></math></span> No significant moderating effects were identified, indicating that mobile learning consistently benefits students across different educational settings. This challenges the belief that mobile learning success depends on context. Instead, the findings underscore its broad potential to enhance learning outcomes in varied environments. This study contributes to theoretical understanding by establishing mobile learning as an effective and adaptable educational strategy and offers practical guidance for educators and policymakers seeking to maximize its benefits.</div></div>","PeriodicalId":10568,"journal":{"name":"Computers & Education","volume":"238 ","pages":"Article 105415"},"PeriodicalIF":8.9,"publicationDate":"2025-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144695458","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"教育学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"K-12 teachers’ ethical competencies for AI literacy: Insights from a systematic literature review","authors":"Deepti Tagare ,&nbsp;Tirtha Karki ,&nbsp;Wonjin Yu","doi":"10.1016/j.compedu.2025.105435","DOIUrl":"10.1016/j.compedu.2025.105435","url":null,"abstract":"<div><div>Artificial Intelligence (AI) technology is now an integral part of our daily life and society. AI offers numerous affordances with potential positive and negative impact on society. K-12 educators are responsible to prepare the future generations to harness the benefits of AI ethically, responsibly, safely, and critically. Teachers need sound knowledge, skills, and dispositions to implement ethics-driven AI literacy education. To achieve this goal, there is a pressing need to identify and operationalize the competencies that teachers need to foster AI literacy. We conducted a larger systematic literature review (SLR), which reviewed empirical studies focused on AI literacy published in the last ten years (2014–2024) to explore and consolidate all competencies for K-12 teachers. This paper reports on teachers’ ethical competencies identified from this SLR. The review of 76 papers included in the final dataset unveils that the ethical AI competencies are interdependent and co-occur in a given context such that teachers may need a combination of knowledge, skills, and dispositions to successfully implement AI literacy education. The findings revealed that teachers need (1) the knowledge competencies of AI ethics issues, limitations of AI, socio-political awareness, and human in the loop, (2) skills of AI transparency and explainability, critical analysis of AI, and culturally responsive and inclusive pedagogy, and (3) the dispositions of AI ethics sensitivity, critical attitude towards AI, and being inclusive and culturally responsive. These ethical competencies can inform teacher preparation and professional development programs for teachers to leverage AI technologies ethically in education.</div></div>","PeriodicalId":10568,"journal":{"name":"Computers & Education","volume":"239 ","pages":"Article 105435"},"PeriodicalIF":10.5,"publicationDate":"2025-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144861287","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"教育学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Digital media use and availability can mitigate the digital divide: Meta-analytic insights from PISA 2022","authors":"Tamara Kastorff,&nbsp;Maren Müller,&nbsp;Samuel Greiff","doi":"10.1016/j.compedu.2025.105409","DOIUrl":"10.1016/j.compedu.2025.105409","url":null,"abstract":"<div><div>The digital divide, commonly described as the relationship between socioeconomic status (SES) and digital competence, has garnered significant attention in educational research, particularly since the onset of the COVID-19 pandemic. However, despite ongoing efforts to promote digitalization across countries and schools, there has been limited research into the extent to which school-level or country-level factors can moderate the relationship between socioeconomic status (SES) and digital competence—an association commonly referred to in the literature as the second-level digital divide. This study addressed these gaps by examining the association between SES and digital competence using data from the Programme for International Student Assessment (PISA) 2022, encompassing 52 countries (<em>N</em> = 393607) and employing an integrative meta-analytical approach. We found positive associations between SES and two measures of digital competencies; self-efficacies, and practices related to online information (<em>β</em>_{summary effect} = 0.15, 95 % CI [0.14, 0.16]. <em>β</em>_{summary effect} = 0.11, 95 % CI [0.10, 0.12]). Furthermore, moderator analyses showed that the adequate availability and high-quality use of digital media, such as inquiry-based learning methods, can mitigate the positive relationship between SES and digital competence, an effect that holds promise for reducing digital inequalities. This study contributes to the existing body of research by providing a more nuanced understanding country-level factors that may mitigate the digital divide. Furthermore, it highlights potential approaches for targeted interventions aimed at addressing SES-related disparities that perpetuate digital inequalities.</div></div>","PeriodicalId":10568,"journal":{"name":"Computers & Education","volume":"238 ","pages":"Article 105409"},"PeriodicalIF":8.9,"publicationDate":"2025-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144613368","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"教育学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}