Giovanni M. Troiano , Amir Abdollahi , Michael Cassidy , Gillian Puttick , Tiago Machado , Casper Harteveld
{"title":"Leveling the computational playing field: Inquiring about factors predicting computational thinking in constructionist game-based learning","authors":"Giovanni M. Troiano , Amir Abdollahi , Michael Cassidy , Gillian Puttick , Tiago Machado , Casper Harteveld","doi":"10.1016/j.compedu.2025.105347","DOIUrl":"10.1016/j.compedu.2025.105347","url":null,"abstract":"<div><div>Computational thinking (CT) is key in STEM and computer science (CS) education. Recently, there has been a surge in studies inquiring about the factors that predict the CT development of young students. We extend these prior works by inquiring about the factors that predict the CT of students (<em>n</em> <span><math><mo>=</mo></math></span> 932) in a constructionist game-based learning (GBL) STEM curriculum. Specifically, after addressing missing data through imputation, we apply Multilevel Modeling (MLM) to identify these potential factors in Scratch games and students’ CT. We found that teachers’ experience implementing game-based curricula, students’ Scratch experience, student choice of game genre, and the interaction between teacher experience and game genre significantly predicted CT. Instead, students’ gender did not emerge as a significant predictor of CT. We provide recommendations for curricula that support CT through constructionist GBL.</div></div>","PeriodicalId":10568,"journal":{"name":"Computers & Education","volume":"237 ","pages":"Article 105347"},"PeriodicalIF":8.9,"publicationDate":"2025-06-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144262358","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":"Enhancing blended learning discussions with a Scaffolded Knowledge Integration–Based ChatGPT mobile instant messaging system","authors":"Hsin-Yu Lee, Ting-Ting Wu","doi":"10.1016/j.compedu.2025.105375","DOIUrl":"10.1016/j.compedu.2025.105375","url":null,"abstract":"<div><div>Recent expansions in the use of ChatGPT within mobile instant messaging (MIM) platforms have garnered attention for their potential to enrich blended learning discussions. However, existing implementations often prioritize quick answers rather than pedagogically structured scaffolds, potentially limiting deeper learning. In this study, we introduce SKIMIM (Scaffolded Knowledge Integration–Based ChatGPT Mobile Instant Messaging), a system designed to systematically incorporate Scaffolded Knowledge Integration framework into AI-supported discussions. SKIMIM prompts learners to elicit their initial ideas, add new concepts, distinguish among different points of view, and reflect to refine their understanding. A 15-week randomized controlled trial (RCT) was conducted with 87 master's students assigned to three groups: SKIMIM, standard ChatGPT-MIM, and traditional MIM. Data were collected through engagement questionnaires, discussion logs, and semi-structured interviews, and analyzed via a mixed-methods approach covering behavioral, cognitive, and emotional dimensions of engagement, as well as user perceptions grounded in an extended Technology Acceptance Model. The results revealed that while standard ChatGPT-MIM provided higher behavioral participation and emotional comfort through rapid AI assistance, SKIMIM significantly enhanced cognitive engagement—particularly by fostering sense-making and innovation-level thinking. Although students experienced an initial adjustment period with SKIMIM's structured prompts, they ultimately reported comparable behavioral intention to use, along with notably higher perceived learning effectiveness and discussion quality. These findings underscore the importance of integrating AI with deliberate scaffolding strategies to achieve both active engagement and deeper cognitive outcomes in blended learning discussions.</div></div>","PeriodicalId":10568,"journal":{"name":"Computers & Education","volume":"237 ","pages":"Article 105375"},"PeriodicalIF":8.9,"publicationDate":"2025-05-31","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144195706","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}
Ahmad Ari Aldino , Yi-Shan Tsai , Siddarth Gupte , Michael Henderson , Debarshi Nath , Dragan Gašević , Guanliang Chen
{"title":"Analytics of Learner-Centered Feedback: A Large-Scale Case Study in Higher Education","authors":"Ahmad Ari Aldino , Yi-Shan Tsai , Siddarth Gupte , Michael Henderson , Debarshi Nath , Dragan Gašević , Guanliang Chen","doi":"10.1016/j.compedu.2025.105360","DOIUrl":"10.1016/j.compedu.2025.105360","url":null,"abstract":"<div><div>Feedback plays a crucial role in guiding students towards achieving their learning goals. The conceptualization of feedback has shifted from <em>teacher-centered</em> to <em>learner-centered</em> approaches, underscoring the evolving role of educators and students in educational settings. Despite the growing emphasis on learner-centered feedback frameworks, there remains a gap in understanding how these frameworks are implemented in actual teaching practices. This case study addresses this gap by examining the alignment of current feedback practices with learner-centered feedback principles in the Computer Science School at an Australian higher education. We gathered feedback data from the Master of Data Science and Bachelor of Computer Science program that were communicated through the Learning Management System. The dataset included feedback from 4959 students, provided by approximately 200 instructors across 95 courses. To ensure a representative sample, 10% of feedback entries from each course were analyzed, resulting in 16,408 feedback sentences. The findings reveal a pronounced emphasis on the sensemaking dimension, particularly in evaluating students’ strengths and weaknesses to help them understand their performance. Feedback patterns varied by student performance, with high achievers receiving affirmations, medium achievers receiving actionable suggestions, and low achievers receiving comprehensive evaluations. Feedback in the Master’s program prioritized future impact by offering actionable guidance for advanced tasks, while the Bachelor’s program emphasized fostering agency through active student engagement and participation.</div></div>","PeriodicalId":10568,"journal":{"name":"Computers & Education","volume":"237 ","pages":"Article 105360"},"PeriodicalIF":8.9,"publicationDate":"2025-05-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144212750","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}
Conrad Borchers , Hendrik Fleischer , Sascha Schanze , Katharina Scheiter , Vincent Aleven
{"title":"High scaffolding of an unfamiliar strategy improves conceptual learning but reduces enjoyment compared to low scaffolding and strategy freedom","authors":"Conrad Borchers , Hendrik Fleischer , Sascha Schanze , Katharina Scheiter , Vincent Aleven","doi":"10.1016/j.compedu.2025.105364","DOIUrl":"10.1016/j.compedu.2025.105364","url":null,"abstract":"<div><div>Adaptive learning systems support students in acquiring complex skills, provided they deliver appropriate instructional support, such as scaffolding. Few studies have examined whether the optimal level of scaffolding depends on the system's support for strategies familiar to the learner—a situation that often arises when students use software developed abroad or aligned with a different curriculum. The present study experimentally compared learning outcomes from two American tutoring systems, StoichTutor and ORCCA, which provide differing levels of scaffolding, in a German population. Prior research predicts that learners would benefit more from a system with flexible support of their native problem-solving approach but provides less scaffolding. To test this prediction, we conducted a crossover experiment involving 61 German undergraduates enrolled in remedial first-year university chemistry preparatory courses. Procedural and conceptual learning were evaluated alongside self-efficacy and usability perceptions. Both tutoring systems significantly promoted procedural learning. However, only the highly scaffolded tutoring system yielded significant conceptual learning gains. Log data analysis revealed that the highly scaffolded system provided more opportunities for students to practice and receive feedback on unit analysis and substance operations. These opportunities corresponded to condition-specific learning gain differences in the underlying conceptual skills. Students significantly preferred working with the highly scaffolded system. These findings suggest that highly scaffolded systems can improve conceptual understanding, even when learners are unfamiliar with the scaffolded strategy. The practical significance of this finding is that, adapting tutoring systems outside of the United States, language translation may suffice to benefit novice learners through scaffolded, adaptive instruction.</div></div>","PeriodicalId":10568,"journal":{"name":"Computers & Education","volume":"236 ","pages":"Article 105364"},"PeriodicalIF":8.9,"publicationDate":"2025-05-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144139167","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":"When and how learners engage with source information in digital multiple-text reading: Effects of task instruction and text trustworthiness from eye-tracking technology","authors":"Zheng-Hong Guan, Sunny S.J. Lin","doi":"10.1016/j.compedu.2025.105362","DOIUrl":"10.1016/j.compedu.2025.105362","url":null,"abstract":"<div><div>Digital reading with multiple-text comprehension is essential in daily life, especially with the rise of AI-generated contents, making source evaluation crucial for authenticity of online information, particularly in academic contexts. Despite its importance, there is limited moment-to-moment evidence on reading processes in multiple-text use. To bridge this gap, we used eye-tracking technology to see how task instructions affect reading multiple texts with high and low trustworthiness. The transitions among texts and source information, as well as how much attention the readers paid are particularly of concern. Sixty-one college students were randomly divided into summary and argument groups, tasked with writing a summary or an argument, respectively. Both groups read four texts presenting conflicting views on genetically modified technology. Each text featured high or low trustworthiness and included three paragraphs: source, evidence, and conclusion. Results of eye-movement data showed both groups spent more time rereading high-trustworthiness texts and evidence paragraphs. The summary group notably reread more on source information, made more transitions between source and evidence, and accessed texts more times than the argument group. In essay writing, the summary group showed better integration by using more source citations and connectives and referring to more concepts, compared to the argument group. Mediation analysis revealed that the process data of rereading time and number of click times successfully mediated the relationship between task instructions and multiple-text integration, confirming that the summary group engaged deeply. These findings offer both theoretical contributions and practical implications for developing adaptive, personalized digital reading environments.</div></div>","PeriodicalId":10568,"journal":{"name":"Computers & Education","volume":"236 ","pages":"Article 105362"},"PeriodicalIF":8.9,"publicationDate":"2025-05-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144116553","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}
Xiaodong Wei , Lei Wang , Tiffany A. Koszalka , Lap-Kei Lee , Ruixue Liu
{"title":"Enhancing pre-service teachers' reflective thinking skills through generative AI-assisted digital storytelling creation: A three-dimensional framework analysis","authors":"Xiaodong Wei , Lei Wang , Tiffany A. Koszalka , Lap-Kei Lee , Ruixue Liu","doi":"10.1016/j.compedu.2025.105356","DOIUrl":"10.1016/j.compedu.2025.105356","url":null,"abstract":"<div><div>Developing reflective thinking skills (RTS) in pre-service teachers remains a challenge in teacher education, particularly in the context of integrating emerging technologies. While digital storytelling (DST) has shown promise in fostering reflective practice, traditional methods often present technical barriers that hinder deeper reflection. Few studies have explored how generative artificial intelligence (GAI) tools can support RTS during DST creation. This study addresses these gaps by adopting a three-dimensional framework of RTS, which included the time of reflection, objects of reflection, and levels of reflection, to guide and assess the impact of GAI-assisted DST creation on pre-service teachers' reflective patterns. Employing a post-test quasi-experimental design, eighty pre-service teachers were divided into two groups: an experimental group utilizing GAI tools (e.g., ChatGPT, Midjourney, Runway) for DST creation, and a control group utilizing traditional methods. Results revealed that the experimental group significantly improved RTS in time, objects, and levels of reflection. Pre-service teachers in the experimental group reflected more on problem definition and solution generation during the design stages of reflection time. Regarding reflection objects, the experimental group exhibited significantly higher reflection frequencies across self, artifacts, and circumstances aspects. Additionally, pre-service teachers in experimental group demonstrated significantly higher reflection frequencies at all levels—single-loop, double-loop, and triple-loop—compared to the control group. Single-loop reflection was the most common, while triple-loop reflection was the least frequent in both groups. These findings underline GAI's potential to scaffold RTS and enhance the DST creation process, offering valuable insights for integrating GAI into teacher education to foster deeper reflective practice.</div></div>","PeriodicalId":10568,"journal":{"name":"Computers & Education","volume":"235 ","pages":"Article 105356"},"PeriodicalIF":8.9,"publicationDate":"2025-05-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144089971","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":"When technology meets emotions: Exploring preschool teachers' emotion profiles in technology use","authors":"Jing Li , Barry Bai , Hao Liu","doi":"10.1016/j.compedu.2025.105355","DOIUrl":"10.1016/j.compedu.2025.105355","url":null,"abstract":"<div><div>The potential of technology in early childhood education (ECE) has been widely recognized. However, limited research has focused on preschool teachers' emotions in technology use and their relationship with emotional intelligence, technological self-efficacy, and school-level variations. To address this gap, this study adopted a multilevel latent profile analysis (MLPA) to explore (1) emotion profiles in technology use, (2) the predictive role of emotional intelligence in profile memberships, and (3) the association between emotion profiles and technological self-efficacy at both the teacher and school levels. Data was collected from 3195 preschool teachers across 268 schools in Shenzhen, China. Specifically, MLPA identified four teacher-level profiles (Suppressed, Balanced, Engaged and Conflicted) and three school-level profiles (Innovative, Steady and Passive), emphasizing substantial variations across schools. Moreover, emotional intelligence, particularly perceiving, using and understanding emotions, predicted memberships in better emotion profiles. The teachers in the better emotion profiles demonstrated higher technological self-efficacy, and school-level differences influenced these relationships. These findings highlight the importance of considering both the individual-level and school-level emotion patterns in supporting effective technology usage within the Chinese cultural context.</div></div>","PeriodicalId":10568,"journal":{"name":"Computers & Education","volume":"236 ","pages":"Article 105355"},"PeriodicalIF":8.9,"publicationDate":"2025-05-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144184413","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":"The relationship between social inhibitions and various measures of communication skills in two types of digital simulations","authors":"Julia Fecke , Katharina Lohberger , Edith Braun","doi":"10.1016/j.compedu.2025.105361","DOIUrl":"10.1016/j.compedu.2025.105361","url":null,"abstract":"<div><div>Teacher education simulations are useful for learning communication skills under authentic conditions. However, they can also inspire social inhibitions, particularly among individuals who experience anxiety in interpersonal situations. Digital simulations have the potential to provide alternative, safe spaces for socially inhibited students to practice communication skills. It is unclear, however, which formats of digital simulations are the most effective. In this study, we analysed video camera-based and avatar-based learning environments in which simulations were conducted. We also used observations and self-reports—before and after the simulation—to measure competences. In an experimental design, sixty-one pre-service teachers were randomly assigned to video camera or avatar-based simulations. Our findings suggest that avatar-based simulations help reduce social inhibitions; students feel more competent after participating in role-plays compared to before. Such effects did not appear in video camera-based simulations, either in self-assessment or external assessment. Due to the extensive time and labour required to conduct this study, the sample was relatively small. These are, therefore, preliminary results that need to be verified in further studies.</div></div>","PeriodicalId":10568,"journal":{"name":"Computers & Education","volume":"235 ","pages":"Article 105361"},"PeriodicalIF":8.9,"publicationDate":"2025-05-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143935431","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":"Bridging the knowledge-skill gap: The role of large language model and critical thinking in education","authors":"Jin Yuxian","doi":"10.1016/j.compedu.2025.105357","DOIUrl":"10.1016/j.compedu.2025.105357","url":null,"abstract":"<div><div>In the dynamic field of education, addressing the knowledge-skill gap remains a significant challenge, as students often excel in theoretical understanding but struggle with practical application. This study investigates the combined effects of the large language model (LLM) chatbot and critical thinking guidance on learners' acquisition of declarative knowledge (“know what”) and procedural knowledge (“know how”). Findings indicate that while the LLM chatbot enhances declarative knowledge acquisition, it does not significantly impact procedural learning, as learners tend to prioritize lower cognitive load and focus on declarative knowledge. In contrast, critical thinking guidance fosters procedural learning but increases cognitive load, thereby limiting resources available for declarative learning. However, when both interventions are combined, they generate synergistic effects—critical thinking guidance activates procedural learning, while the LLM chatbot mitigates cognitive burden, enabling a more balanced allocation of cognitive resources and improving both declarative and procedural knowledge acquisition. These findings highlight the potential of LLM chatbots as effective educational tools, suggesting that the strategic use of artificial intelligence can promote a more effective approach to knowledge acquisition. This research provides valuable insights into the application of advanced technologies in educational contexts, emphasizing the importance of appropriate instructional strategies to guide the effective use of these technologies.</div></div>","PeriodicalId":10568,"journal":{"name":"Computers & Education","volume":"235 ","pages":"Article 105357"},"PeriodicalIF":8.9,"publicationDate":"2025-05-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143941156","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":"Development and validation of moral ascendancy and dependency in AI integration (MAD-AI) scale for teachers","authors":"Joanne Jorolan , Florejane Cabillo , Renna Rose Batucan , Cherish Mae Camansi , Angela Etoquilla , Jessieca Gapo , Danica Kaye Hallarte , Masza Lyn Milano , Roselyn Gonzales , Gamaliel Gonzales","doi":"10.1016/j.compedu.2025.105346","DOIUrl":"10.1016/j.compedu.2025.105346","url":null,"abstract":"<div><div>Integrating artificial intelligence into teaching practices has become a prominent topic in recent literature, raising concerns about how teachers' moral ascendancy and dependency can be attributed and measured in a behavioral context. This study addresses this gap by developing and validating the teachers' perceived moral ascendancy and dependency on the AI integration (MAD-AI) scale. The methodological process involved the following key steps: initially identifying the scale's dimensionality and creating draft items through a literature review; conducting validation with four experts; revising items based on expert feedback; holding two focus group discussions with six preservice and four in-service teachers; refining dimensions and items based on these discussions; and finally, establishing the psychometric properties of the scale through exploratory factor analysis and confirmatory factor analysis with 383 survey responses from 224 in-service and 159 preservice teachers in the Philippines. The final 30-item MAD-AI scale for teachers includes two dimensions for moral ascendancy: (a) ethical transparency and accountability (7 items) and (b) professional integrity (8 items), as well as two dimensions for dependency: (a) institutional support (8 items) and (b) educator preparedness (7 items). Results demonstrated acceptable fit measures, strong reliability, convergent validity, and discriminant validity, supporting the structural soundness of the scale.</div></div>","PeriodicalId":10568,"journal":{"name":"Computers & Education","volume":"235 ","pages":"Article 105346"},"PeriodicalIF":8.9,"publicationDate":"2025-05-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143924511","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}