Collaborative problem solving (CPS) is an important skill enabling students to co-construct knowledge and tackle complex problems through group interactions. While the importance of group interactions in CPS is well recognized, it is unclear how the emergence of generative artificial intelligence (GAI), with advanced cognitive support, may alter group dynamics in CPS. This study bridges this gap by examining group interactions in GAI-supported CPS, focusing on the structural patterns and interaction content characterizing students' social dynamics. Six groups of three to five students used an online messaging tool with a GPT-4.0 enabled chatbot for a CPS activity. Group interactions were modelled using network analysis and interaction content was coded into socio-emotional, cognitive, metacognitive, and coordinative dimensions. Employing a network assortativity measure and a binomial test to the interactions among students and the GAI chatbot, we identified a GAI-centred interaction pattern in which students tended to interact significantly more with the chatbot than their peers in the collaborative problem-solving process. Students' interactions with the chatbot involved primarily cognitive interactions but also metacognitive and socio-emotional interactions. This study introduces novel network methods to analyse small group interactions and contributes new empirical evidence and theoretical insights into the social influence of GAI tools, emphasizing the need for further investigations on the factors influencing interaction dynamics among students and GAI tools in collaborative learning.
{"title":"Group interaction patterns in generative AI-supported collaborative problem solving: Network analysis of the interactions among students and a GAI chatbot","authors":"Shihui Feng","doi":"10.1111/bjet.13611","DOIUrl":"10.1111/bjet.13611","url":null,"abstract":"<p>Collaborative problem solving (CPS) is an important skill enabling students to co-construct knowledge and tackle complex problems through group interactions. While the importance of group interactions in CPS is well recognized, it is unclear how the emergence of generative artificial intelligence (GAI), with advanced cognitive support, may alter group dynamics in CPS. This study bridges this gap by examining group interactions in GAI-supported CPS, focusing on the structural patterns and interaction content characterizing students' social dynamics. Six groups of three to five students used an online messaging tool with a GPT-4.0 enabled chatbot for a CPS activity. Group interactions were modelled using network analysis and interaction content was coded into socio-emotional, cognitive, metacognitive, and coordinative dimensions. Employing a network assortativity measure and a binomial test to the interactions among students and the GAI chatbot, we identified a GAI-centred interaction pattern in which students tended to interact significantly more with the chatbot than their peers in the collaborative problem-solving process. Students' interactions with the chatbot involved primarily cognitive interactions but also metacognitive and socio-emotional interactions. This study introduces novel network methods to analyse small group interactions and contributes new empirical evidence and theoretical insights into the social influence of GAI tools, emphasizing the need for further investigations on the factors influencing interaction dynamics among students and GAI tools in collaborative learning.</p>","PeriodicalId":48315,"journal":{"name":"British Journal of Educational Technology","volume":"56 5","pages":"2125-2145"},"PeriodicalIF":8.1,"publicationDate":"2025-06-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://bera-journals.onlinelibrary.wiley.com/doi/epdf/10.1111/bjet.13611","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144809309","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"教育学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Shahper Richter, Shohil Kishore, Inna Piven, Patrick Dodd, Guy Bate
<div> <section> <p>This study investigates how anthropomorphic AI chatbot avatars, designed in line with the Stereotype Content Model (SCM) dimensions of warmth and competence, influence university students' perceptions of support for self-directed learning (SDL) activities. We examined student responses to two distinct avatars—one projecting warmth and the other projecting competence. Using an Action Design Research (ADR) approach, we evaluated the chatbots across three university courses, incorporating perspectives from students, educators and learning designers. Findings reveal that students perceive the avatars differently. The warm avatar provides a stronger emotional connection, while the competent avatar offers more effective task-oriented learning support. These results highlight the importance of balancing warmth and competence in chatbot design to enhance their perceived usefulness for supporting SDL engagement. The study also supplies rich insights into practical implementation challenges and opportunities from multiple stakeholder viewpoints. Altogether, the research advances our understanding of SCM-informed chatbot design in educational settings and proposes practical principles for developing AI tools that students perceive as helpful, thereby contributing to the field of human–AI interaction.</p> </section> <section> <div> <div> <h3>Practitioner notes</h3> <p>What is already known about this topic </p><ul> <li>The potential of AI chatbots to support aspects of self-directed learning (SDL) in higher education is currently being explored.</li> <li>User perceptions of AI systems are influenced by anthropomorphic design cues, often understood through dimensions like warmth and competence (related to the Stereotype Content Model—SCM).</li> <li>Designing AI educational tools requires considering how different interactional styles (eg, warmth vs. competence) can affect student engagement and perceived usefulness.</li> </ul> <p>What this paper adds </p><ul> <li>Empirical insights into students' perceptions of chatbot avatars designed with varying levels of warmth and competence, based on the SCM, and how these perceptions relate to their reported engagement and perceived support for SDL in university courses.</li> <li>Evidence that students distinguish between warmth and competence in chatbot avatars, associating warmth with socio-emotional connection and competence with task-related lea
{"title":"Chatbots in tertiary education: Exploring the impact of warm and competent avatars on self-directed learning","authors":"Shahper Richter, Shohil Kishore, Inna Piven, Patrick Dodd, Guy Bate","doi":"10.1111/bjet.13610","DOIUrl":"10.1111/bjet.13610","url":null,"abstract":"<div>\u0000 \u0000 <section>\u0000 \u0000 <p>This study investigates how anthropomorphic AI chatbot avatars, designed in line with the Stereotype Content Model (SCM) dimensions of warmth and competence, influence university students' perceptions of support for self-directed learning (SDL) activities. We examined student responses to two distinct avatars—one projecting warmth and the other projecting competence. Using an Action Design Research (ADR) approach, we evaluated the chatbots across three university courses, incorporating perspectives from students, educators and learning designers. Findings reveal that students perceive the avatars differently. The warm avatar provides a stronger emotional connection, while the competent avatar offers more effective task-oriented learning support. These results highlight the importance of balancing warmth and competence in chatbot design to enhance their perceived usefulness for supporting SDL engagement. The study also supplies rich insights into practical implementation challenges and opportunities from multiple stakeholder viewpoints. Altogether, the research advances our understanding of SCM-informed chatbot design in educational settings and proposes practical principles for developing AI tools that students perceive as helpful, thereby contributing to the field of human–AI interaction.</p>\u0000 </section>\u0000 \u0000 <section>\u0000 \u0000 <div>\u0000 \u0000 <div>\u0000 \u0000 <h3>Practitioner notes</h3>\u0000 <p>What is already known about this topic\u0000\u0000 </p><ul>\u0000 \u0000 <li>The potential of AI chatbots to support aspects of self-directed learning (SDL) in higher education is currently being explored.</li>\u0000 \u0000 <li>User perceptions of AI systems are influenced by anthropomorphic design cues, often understood through dimensions like warmth and competence (related to the Stereotype Content Model—SCM).</li>\u0000 \u0000 <li>Designing AI educational tools requires considering how different interactional styles (eg, warmth vs. competence) can affect student engagement and perceived usefulness.</li>\u0000 </ul>\u0000 <p>What this paper adds\u0000\u0000 </p><ul>\u0000 \u0000 <li>Empirical insights into students' perceptions of chatbot avatars designed with varying levels of warmth and competence, based on the SCM, and how these perceptions relate to their reported engagement and perceived support for SDL in university courses.</li>\u0000 \u0000 <li>Evidence that students distinguish between warmth and competence in chatbot avatars, associating warmth with socio-emotional connection and competence with task-related lea","PeriodicalId":48315,"journal":{"name":"British Journal of Educational Technology","volume":"56 5","pages":"2102-2124"},"PeriodicalIF":8.1,"publicationDate":"2025-06-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://bera-journals.onlinelibrary.wiley.com/doi/epdf/10.1111/bjet.13610","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144811105","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"教育学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Sandra Fischer-Schöneborn, Chris Brown, Burak Aydin, Stephen MacGregor, Marcus Pietsch
<div> <section> <p>This study examines whether and how a school's information and communication technology (ICT) knowledge absorptive capacity (ACAP) affects technology integration in schools. In addition, it investigates the influence of various contextual factors on the degree of contingency of ACAP, such as activation triggers, social integration mechanisms and regimes of appropriability. The study is based on a random sample of <i>N</i> = 411 schools representative of Germany. Structural equation modelling and machine learning were employed. The findings indicate that ICT ACAP has a positive impact on technology integration in schools and serves as a mediator in the relationship between external knowledge and technology integration. The impact of ICT ACAP on technology integration is contingent upon the presence and efficacy of knowledge-sharing mechanisms within the school, as well as the extent to which schools engage in collaborative efforts with competitors (coopetition). The insights of this study have implications for policymakers and educational leaders, who could prioritize building ACAP and fostering collaborative networks to create more adaptable and innovative school environments.</p> </section> <section> <div> <div> <h3>Practitioner notes</h3> <p>What is already known about this topic </p><ul> <li>For schools, technology integration is considered an important educational innovation.</li> <li>Acquiring, creating and sharing knowledge are essential for an efficient technology integration.</li> <li>Knowledge absorptive capacity (ACAP) is a critical factor in the acquisition of knowledge.</li> </ul> <p>What this paper adds </p><ul> <li>Higher information and communication technology (ICT) ACAP is associated with increased technology integration.</li> <li>ICT ACAP mediates between the depth of external knowledge and technology integration.</li> <li>The efficacy of ACAP is contingent upon a number of contextual variables, in particular, knowledge sharing in schools and coopetition.</li> </ul> <p>Implications for practice and/or policy </p><ul> <li>Schools need to identify, integrate and exploit relevant ICT knowledge to integrate technology successfully.</li> <li
{"title":"ICT knowledge absorptive capacity: A critical factor for technology integration in schools","authors":"Sandra Fischer-Schöneborn, Chris Brown, Burak Aydin, Stephen MacGregor, Marcus Pietsch","doi":"10.1111/bjet.13608","DOIUrl":"https://doi.org/10.1111/bjet.13608","url":null,"abstract":"<div>\u0000 \u0000 \u0000 <section>\u0000 \u0000 <p>This study examines whether and how a school's information and communication technology (ICT) knowledge absorptive capacity (ACAP) affects technology integration in schools. In addition, it investigates the influence of various contextual factors on the degree of contingency of ACAP, such as activation triggers, social integration mechanisms and regimes of appropriability. The study is based on a random sample of <i>N</i> = 411 schools representative of Germany. Structural equation modelling and machine learning were employed. The findings indicate that ICT ACAP has a positive impact on technology integration in schools and serves as a mediator in the relationship between external knowledge and technology integration. The impact of ICT ACAP on technology integration is contingent upon the presence and efficacy of knowledge-sharing mechanisms within the school, as well as the extent to which schools engage in collaborative efforts with competitors (coopetition). The insights of this study have implications for policymakers and educational leaders, who could prioritize building ACAP and fostering collaborative networks to create more adaptable and innovative school environments.</p>\u0000 </section>\u0000 \u0000 <section>\u0000 \u0000 <div>\u0000 \u0000 <div>\u0000 \u0000 <h3>Practitioner notes</h3>\u0000 <p>What is already known about this topic\u0000\u0000 </p><ul>\u0000 \u0000 <li>For schools, technology integration is considered an important educational innovation.</li>\u0000 \u0000 <li>Acquiring, creating and sharing knowledge are essential for an efficient technology integration.</li>\u0000 \u0000 <li>Knowledge absorptive capacity (ACAP) is a critical factor in the acquisition of knowledge.</li>\u0000 </ul>\u0000 <p>What this paper adds\u0000\u0000 </p><ul>\u0000 \u0000 <li>Higher information and communication technology (ICT) ACAP is associated with increased technology integration.</li>\u0000 \u0000 <li>ICT ACAP mediates between the depth of external knowledge and technology integration.</li>\u0000 \u0000 <li>The efficacy of ACAP is contingent upon a number of contextual variables, in particular, knowledge sharing in schools and coopetition.</li>\u0000 </ul>\u0000 <p>Implications for practice and/or policy\u0000\u0000 </p><ul>\u0000 \u0000 <li>Schools need to identify, integrate and exploit relevant ICT knowledge to integrate technology successfully.</li>\u0000 \u0000 <li","PeriodicalId":48315,"journal":{"name":"British Journal of Educational Technology","volume":"57 1","pages":"138-162"},"PeriodicalIF":8.1,"publicationDate":"2025-06-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://bera-journals.onlinelibrary.wiley.com/doi/epdf/10.1111/bjet.13608","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145969920","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"教育学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Belle Dang, Luna Huynh, Faaiz Gul, Carolyn Rosé, Sanna Järvelä, Andy Nguyen
<div> <section> <p>The rise of generative artificial intelligence (GAI), especially with multimodal large language models like GPT-4o, sparked transformative potential and challenges for learning and teaching. With potential as a cognitive offloading tool, GAI can enable learners to focus on higher-order thinking and creativity. Yet, this also raises questions about integration into traditional education due to the limited research on learners' interactions with GAI. Some studies with GAI focus on text-based human–AI interactions, while research on embodied GAI in immersive environments like mixed reality (MR) remains unexplored. To address this, this study investigates interaction dynamics between learners and embodied GAI agents in MR, examining cognitive and socio-emotional interactions during collaborative learning. We investigated the paired interactive patterns between a student and an embodied GAI agent in MR, based on data from 26 higher education students with 1317 recorded activities. Data were analysed using a multi-layered learning analytics approach, including quantitative content analysis, sequence analysis via hierarchical clustering and pattern analysis through ordered network analysis (ONA). Our findings identified two interaction patterns: type (1) AI-led Supported Exploratory Questioning (AISQ) and type (2) Learner-Initiated Inquiry (LII) group. Despite their distinction in characteristic, both types demonstrated comparable levels of socio-emotional engagement and exhibited meaningful cognitive engagement, surpassing the superficial content reproduction that can be observed in interactions with GPT models. This study contributes to the human–AI collaboration and learning studies, extending understanding to learning in MR environments and highlighting implications for designing AI-based educational tools.</p> </section> <section> <div> <div> <h3>Practitioner notes</h3> <p>What is already known about this topic </p><ul> <li>Socio-emotional interactions are fundamental to cognitive processes and play a critical role in collaborative learning.</li> <li>Generative artificial intelligence (GAI) holds transformative potential for education but raises questions about how learners interact with such technology.</li> <li>Most existing research focuses on text-based interactions with GAI; there is limited empirical evidence on how embodied GAI agents within immersive environments like Mixed Reality (MR) influence the cognitive and socio-emotional interactions for learning and regulation.</li> </ul> <p>Wha
{"title":"Human–AI collaborative learning in mixed reality: Examining the cognitive and socio-emotional interactions","authors":"Belle Dang, Luna Huynh, Faaiz Gul, Carolyn Rosé, Sanna Järvelä, Andy Nguyen","doi":"10.1111/bjet.13607","DOIUrl":"10.1111/bjet.13607","url":null,"abstract":"<div>\u0000 \u0000 <section>\u0000 \u0000 <p>The rise of generative artificial intelligence (GAI), especially with multimodal large language models like GPT-4o, sparked transformative potential and challenges for learning and teaching. With potential as a cognitive offloading tool, GAI can enable learners to focus on higher-order thinking and creativity. Yet, this also raises questions about integration into traditional education due to the limited research on learners' interactions with GAI. Some studies with GAI focus on text-based human–AI interactions, while research on embodied GAI in immersive environments like mixed reality (MR) remains unexplored. To address this, this study investigates interaction dynamics between learners and embodied GAI agents in MR, examining cognitive and socio-emotional interactions during collaborative learning. We investigated the paired interactive patterns between a student and an embodied GAI agent in MR, based on data from 26 higher education students with 1317 recorded activities. Data were analysed using a multi-layered learning analytics approach, including quantitative content analysis, sequence analysis via hierarchical clustering and pattern analysis through ordered network analysis (ONA). Our findings identified two interaction patterns: type (1) AI-led Supported Exploratory Questioning (AISQ) and type (2) Learner-Initiated Inquiry (LII) group. Despite their distinction in characteristic, both types demonstrated comparable levels of socio-emotional engagement and exhibited meaningful cognitive engagement, surpassing the superficial content reproduction that can be observed in interactions with GPT models. This study contributes to the human–AI collaboration and learning studies, extending understanding to learning in MR environments and highlighting implications for designing AI-based educational tools.</p>\u0000 </section>\u0000 \u0000 <section>\u0000 \u0000 <div>\u0000 \u0000 <div>\u0000 \u0000 <h3>Practitioner notes</h3>\u0000 <p>What is already known about this topic\u0000\u0000 </p><ul>\u0000 \u0000 <li>Socio-emotional interactions are fundamental to cognitive processes and play a critical role in collaborative learning.</li>\u0000 \u0000 <li>Generative artificial intelligence (GAI) holds transformative potential for education but raises questions about how learners interact with such technology.</li>\u0000 \u0000 <li>Most existing research focuses on text-based interactions with GAI; there is limited empirical evidence on how embodied GAI agents within immersive environments like Mixed Reality (MR) influence the cognitive and socio-emotional interactions for learning and regulation.</li>\u0000 </ul>\u0000 <p>Wha","PeriodicalId":48315,"journal":{"name":"British Journal of Educational Technology","volume":"56 5","pages":"2078-2101"},"PeriodicalIF":8.1,"publicationDate":"2025-06-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://bera-journals.onlinelibrary.wiley.com/doi/epdf/10.1111/bjet.13607","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144811294","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"教育学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
<div> <section> <p>ChatGPT is a generative Artificial Intelligence (AI) that can produce a variety of outputs, including solutions to problems. Prior research shows that for students to learn from instructional content, they need to actively process the content. To date, existing research has focused on student explanations expressed in words (either spoken or written). Thus, less is known about other forms of expression, such as ones involving spatial elements (eg, flowcharts, drawings). Moreover, to the best of our knowledge, there is not yet work on how students annotate solutions produced by ChatGPT. The study was conducted in the context of a first-year programming tutorial focused on loops and code tracing. Code tracing is a fundamental programming skill that involves simulating at a high level the actions a computer takes when it executes a program. The students annotated a printed-out code trace produced by ChatGPT using the strategy of their choice. Our goal was to describe the visual and verbal strategies students used to annotate the ChatGPT trace as well as how strategies relate to annotation quality, and so we used an observational study design with a single condition. Annotation strategies ranged from words-only strategies to visual representations like flowcharts. As annotation quality increased, the proportions of strategies used changed, suggesting that some strategies may facilitate the production of quality annotations. In particular, the proportion of words-only and flowchart strategies increased as quality increased; in the top quality quartile, there was a similar proportion of each but with slightly more flowcharts.</p> </section> <section> <div> <div> <h3>Practitioner notes</h3> <p>What is already known about this topic </p><ul> <li>When students study instructional materials, they need to actively and constructively interact with the materials in order to learn effectively. Much of the research showing this has examined only verbal student output.</li> <li>In addition to verbal strategies involving only words, strategies including visual elements are also beneficial. For instance, when students are asked to predict a program's output by simulating the steps the computer takes when executing the program, they use representations like tables and/or visual elements to organise their work. These strategies are positively associated with tracing performance.</li> <li>To date, research has focused on how students study instructional materials produced by humans, rather than Large Language Models.</li>
{"title":"Beyond verbal self-explanations: Student annotations of a code-tracing solution produced by ChatGPT","authors":"Abbey Gandhi, Kasia Muldner","doi":"10.1111/bjet.13600","DOIUrl":"https://doi.org/10.1111/bjet.13600","url":null,"abstract":"<div>\u0000 \u0000 \u0000 <section>\u0000 \u0000 <p>ChatGPT is a generative Artificial Intelligence (AI) that can produce a variety of outputs, including solutions to problems. Prior research shows that for students to learn from instructional content, they need to actively process the content. To date, existing research has focused on student explanations expressed in words (either spoken or written). Thus, less is known about other forms of expression, such as ones involving spatial elements (eg, flowcharts, drawings). Moreover, to the best of our knowledge, there is not yet work on how students annotate solutions produced by ChatGPT. The study was conducted in the context of a first-year programming tutorial focused on loops and code tracing. Code tracing is a fundamental programming skill that involves simulating at a high level the actions a computer takes when it executes a program. The students annotated a printed-out code trace produced by ChatGPT using the strategy of their choice. Our goal was to describe the visual and verbal strategies students used to annotate the ChatGPT trace as well as how strategies relate to annotation quality, and so we used an observational study design with a single condition. Annotation strategies ranged from words-only strategies to visual representations like flowcharts. As annotation quality increased, the proportions of strategies used changed, suggesting that some strategies may facilitate the production of quality annotations. In particular, the proportion of words-only and flowchart strategies increased as quality increased; in the top quality quartile, there was a similar proportion of each but with slightly more flowcharts.</p>\u0000 </section>\u0000 \u0000 <section>\u0000 \u0000 <div>\u0000 \u0000 <div>\u0000 \u0000 <h3>Practitioner notes</h3>\u0000 <p>What is already known about this topic\u0000\u0000 </p><ul>\u0000 \u0000 <li>When students study instructional materials, they need to actively and constructively interact with the materials in order to learn effectively. Much of the research showing this has examined only verbal student output.</li>\u0000 \u0000 <li>In addition to verbal strategies involving only words, strategies including visual elements are also beneficial. For instance, when students are asked to predict a program's output by simulating the steps the computer takes when executing the program, they use representations like tables and/or visual elements to organise their work. These strategies are positively associated with tracing performance.</li>\u0000 \u0000 <li>To date, research has focused on how students study instructional materials produced by humans, rather than Large Language Models.</li>\u0000 ","PeriodicalId":48315,"journal":{"name":"British Journal of Educational Technology","volume":"57 1","pages":"112-137"},"PeriodicalIF":8.1,"publicationDate":"2025-05-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://bera-journals.onlinelibrary.wiley.com/doi/epdf/10.1111/bjet.13600","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145984000","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"教育学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
ChanMin Kim, Rebecca J. Passonneau, Eunseo Lee, Mahsa Sheikhi Karizaki, Dana Gnesdilow, Sadhana Puntambekar
<div> <section> <p>As use of artificial intelligence (AI) has increased, concerns about AI bias and discrimination have been growing. This paper discusses an application called PyrEval in which natural language processing (NLP) was used to automate assessment and provide feedback on middle school science writing without linguistic discrimination. Linguistic discrimination in this study was operationalized as unfair assessment of scientific essays based on writing features that are not considered normative such as subject-verb disagreement. Such unfair assessment is especially problematic when the purpose of assessment is not assessing English writing but rather assessing the content of scientific explanations. PyrEval was implemented in middle school science classrooms. Students explained their roller coaster design by stating relationships among such science concepts as potential energy, kinetic energy and law of conservation of energy. Initial and revised versions of scientific essays written by 307 eighth-grade students were analyzed. Our manual and NLP assessment comparison analysis showed that PyrEval did not penalize student essays that contained non-normative writing features. Repeated measures ANOVAs and GLMM analysis results revealed that essay quality significantly improved from initial to revised essays after receiving the NLP feedback, regardless of non-normative writing features. Findings and implications are discussed.</p> </section> <section> <div> <div> <h3>Practitioner notes</h3> <p>What is already known about this topic </p><ul> <li>Advancement in AI has created a variety of opportunities in education, including automated assessment, but AI is not bias-free.</li> <li>Automated writing assessment designed to improve students' scientific explanations has been studied.</li> <li>While limited, some studies reported biased performance of automated writing assessment tools, but without looking into actual linguistic features about which the tools may have discriminated.</li> </ul> <p>What this paper adds </p><ul> <li>This study conducted an actual examination of non-normative linguistic features in essays written by middle school students to uncover how our NLP tool called PyrEval worked to assess them.</li> <li>PyrEval did not penalize essays containing non-normative linguistic features.</li> <li>Regardless of non-normati
{"title":"NLP-enabled automated assessment of scientific explanations: Towards eliminating linguistic discrimination","authors":"ChanMin Kim, Rebecca J. Passonneau, Eunseo Lee, Mahsa Sheikhi Karizaki, Dana Gnesdilow, Sadhana Puntambekar","doi":"10.1111/bjet.13596","DOIUrl":"https://doi.org/10.1111/bjet.13596","url":null,"abstract":"<div>\u0000 \u0000 \u0000 <section>\u0000 \u0000 <p>As use of artificial intelligence (AI) has increased, concerns about AI bias and discrimination have been growing. This paper discusses an application called PyrEval in which natural language processing (NLP) was used to automate assessment and provide feedback on middle school science writing without linguistic discrimination. Linguistic discrimination in this study was operationalized as unfair assessment of scientific essays based on writing features that are not considered normative such as subject-verb disagreement. Such unfair assessment is especially problematic when the purpose of assessment is not assessing English writing but rather assessing the content of scientific explanations. PyrEval was implemented in middle school science classrooms. Students explained their roller coaster design by stating relationships among such science concepts as potential energy, kinetic energy and law of conservation of energy. Initial and revised versions of scientific essays written by 307 eighth-grade students were analyzed. Our manual and NLP assessment comparison analysis showed that PyrEval did not penalize student essays that contained non-normative writing features. Repeated measures ANOVAs and GLMM analysis results revealed that essay quality significantly improved from initial to revised essays after receiving the NLP feedback, regardless of non-normative writing features. Findings and implications are discussed.</p>\u0000 </section>\u0000 \u0000 <section>\u0000 \u0000 <div>\u0000 \u0000 <div>\u0000 \u0000 <h3>Practitioner notes</h3>\u0000 <p>What is already known about this topic\u0000\u0000 </p><ul>\u0000 \u0000 <li>Advancement in AI has created a variety of opportunities in education, including automated assessment, but AI is not bias-free.</li>\u0000 \u0000 <li>Automated writing assessment designed to improve students' scientific explanations has been studied.</li>\u0000 \u0000 <li>While limited, some studies reported biased performance of automated writing assessment tools, but without looking into actual linguistic features about which the tools may have discriminated.</li>\u0000 </ul>\u0000 <p>What this paper adds\u0000\u0000 </p><ul>\u0000 \u0000 <li>This study conducted an actual examination of non-normative linguistic features in essays written by middle school students to uncover how our NLP tool called PyrEval worked to assess them.</li>\u0000 \u0000 <li>PyrEval did not penalize essays containing non-normative linguistic features.</li>\u0000 \u0000 <li>Regardless of non-normati","PeriodicalId":48315,"journal":{"name":"British Journal of Educational Technology","volume":"57 1","pages":"79-111"},"PeriodicalIF":8.1,"publicationDate":"2025-05-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://bera-journals.onlinelibrary.wiley.com/doi/epdf/10.1111/bjet.13596","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145983754","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"教育学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Anni Silvola, Anu Kajamaa, Joonas Merikko, Hanni Muukkonen
Despite a proliferation of research on generative artificial intelligence (GenAI) and its applications in higher education (HE), our understanding of the transformative processes where students create productive and ethically grounded uses of GenAI and how AI mediates students' sensemaking is still limited. Based on an empirical investigation of bachelor's degree students from educational sciences (N = 22) carrying out an inquiry-based course assignment, we analysed students' reflective essays to explore how GenAI mediated their sensemaking throughout the academic writing process. We selected an abductive analysis as the main approach to examine the AI-mediated construction of new understanding. Cross-tabulation analysis complemented qualitative analysis, addressing differences in AI-mediated sensemaking processes based on students' age. Our findings capture a multidimensional constellation of AI-mediated sensemaking processes. We found three central dynamics that guided students' sensemaking process: assessing and adapting the textual characteristics of AI-mediated writing, adjusting and improving interactions with GenAI, and contextualising AI-mediated academic writing experiences around everyday study practices. The tensions and ambiguities highlighted the ethical aspects of adopting AI-mediated academic writing practices, although students did not overcome all of these tensions during their sensemaking processes. Our study contributes theoretically by developing the notion of an AI-mediated sensemaking approach, therefore adding to existing understanding about the dialogical trajectories of AI-mediated writing processes through which students create new meanings and understandings of GenAI use as a learning resource. Further, we discuss the collective aspects of AI-mediated sensemaking.
{"title":"AI-mediated sensemaking in higher education students’ learning processes: Tensions, sensemaking practices, and AI-assigned purposes","authors":"Anni Silvola, Anu Kajamaa, Joonas Merikko, Hanni Muukkonen","doi":"10.1111/bjet.13606","DOIUrl":"10.1111/bjet.13606","url":null,"abstract":"<p>Despite a proliferation of research on generative artificial intelligence (GenAI) and its applications in higher education (HE), our understanding of the transformative processes where students create productive and ethically grounded uses of GenAI and how AI mediates students' sensemaking is still limited. Based on an empirical investigation of bachelor's degree students from educational sciences (<i>N</i> = 22) carrying out an inquiry-based course assignment, we analysed students' reflective essays to explore how GenAI mediated their sensemaking throughout the academic writing process. We selected an abductive analysis as the main approach to examine the AI-mediated construction of new understanding. Cross-tabulation analysis complemented qualitative analysis, addressing differences in AI-mediated sensemaking processes based on students' age. Our findings capture a multidimensional constellation of AI-mediated sensemaking processes. We found three central dynamics that guided students' sensemaking process: assessing and adapting the textual characteristics of AI-mediated writing, adjusting and improving interactions with GenAI, and contextualising AI-mediated academic writing experiences around everyday study practices. The tensions and ambiguities highlighted the ethical aspects of adopting AI-mediated academic writing practices, although students did not overcome all of these tensions during their sensemaking processes. Our study contributes theoretically by developing the notion of an AI-mediated sensemaking approach, therefore adding to existing understanding about the dialogical trajectories of AI-mediated writing processes through which students create new meanings and understandings of GenAI use as a learning resource. Further, we discuss the collective aspects of AI-mediated sensemaking.</p>","PeriodicalId":48315,"journal":{"name":"British Journal of Educational Technology","volume":"56 5","pages":"2001-2018"},"PeriodicalIF":8.1,"publicationDate":"2025-05-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://bera-journals.onlinelibrary.wiley.com/doi/epdf/10.1111/bjet.13606","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144811127","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"教育学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
<div> <section> <p>This study examines the impact of feedback assisted by generative artificial intelligence (GAI) on the dynamics and outcomes of online inquiry-based discussions (IBDs) in a higher education context. Specifically, it compares two distinct feedback types powered by GAI: idea-oriented and task-oriented. The study involved 105 preservice teachers from a public university in Northwestern China. Participants were pre-assigned into two classes, each receiving different types of GAI-assisted feedback during IBDs. A collection of data, including discussion transcripts, survey responses, and IBD performance, was collected and analysed with statistical methods and epistemic network analysis. The results demonstrated that idea-oriented feedback significantly enhanced cognitive presence and led to higher engagement in the exploration of different ideas and opinions. However, this type of feedback also induced greater negative emotional responses. In contrast, task-oriented feedback promoted more social interaction and group cohesion, though it was less effective in fostering higher-order thinking. The findings suggest that GAI tools can provide meaningful support in online learning settings, but the type of feedback must be carefully aligned with the desired learning outcomes. This research offers insights for optimizing GAI-assisted feedback mechanisms in higher education.</p> </section> <section> <div> <div> <h3>Practitioner notes</h3> <p>What is already known about this topic </p><ul> <li>Feedback is key to fostering collaborative problem-solving and critical thinking in online inquiry-based discussions (IBDs).</li> <li>The Community of Inquiry (CoI) model emphasizes the interaction of cognitive, social, and teaching presence for worthwhile learning, with feedback playing a crucial role in regulating these presences.</li> <li>Generative artificial intelligence (GAI) tools have shown potential for providing real-time and personalized feedback.</li> </ul> <p>What this paper adds </p><ul> <li>This study examines two types of GAI-assisted feedback (idea-oriented and task-oriented) and their impact on the dynamics and outcomes of online IBDs.</li> <li>Idea-oriented feedback significantly enhances cognitive presence and promotes deeper inquiry, while task-oriented feedback fosters social presence and group cohesion.</li> <li>GAI-assisted feedback, when aligned with specific learning objectives, can meaningfully promote IBD effective
{"title":"Investigating the impact of ChatGPT-assisted feedback on the dynamics and outcomes of online inquiry-based discussion","authors":"Shen Ba, Ying Zhan, Lingyun Huang, Guoqing Lu","doi":"10.1111/bjet.13605","DOIUrl":"10.1111/bjet.13605","url":null,"abstract":"<div>\u0000 \u0000 <section>\u0000 \u0000 <p>This study examines the impact of feedback assisted by generative artificial intelligence (GAI) on the dynamics and outcomes of online inquiry-based discussions (IBDs) in a higher education context. Specifically, it compares two distinct feedback types powered by GAI: idea-oriented and task-oriented. The study involved 105 preservice teachers from a public university in Northwestern China. Participants were pre-assigned into two classes, each receiving different types of GAI-assisted feedback during IBDs. A collection of data, including discussion transcripts, survey responses, and IBD performance, was collected and analysed with statistical methods and epistemic network analysis. The results demonstrated that idea-oriented feedback significantly enhanced cognitive presence and led to higher engagement in the exploration of different ideas and opinions. However, this type of feedback also induced greater negative emotional responses. In contrast, task-oriented feedback promoted more social interaction and group cohesion, though it was less effective in fostering higher-order thinking. The findings suggest that GAI tools can provide meaningful support in online learning settings, but the type of feedback must be carefully aligned with the desired learning outcomes. This research offers insights for optimizing GAI-assisted feedback mechanisms in higher education.</p>\u0000 </section>\u0000 \u0000 <section>\u0000 \u0000 <div>\u0000 \u0000 <div>\u0000 \u0000 <h3>Practitioner notes</h3>\u0000 <p>What is already known about this topic\u0000\u0000 </p><ul>\u0000 \u0000 <li>Feedback is key to fostering collaborative problem-solving and critical thinking in online inquiry-based discussions (IBDs).</li>\u0000 \u0000 <li>The Community of Inquiry (CoI) model emphasizes the interaction of cognitive, social, and teaching presence for worthwhile learning, with feedback playing a crucial role in regulating these presences.</li>\u0000 \u0000 <li>Generative artificial intelligence (GAI) tools have shown potential for providing real-time and personalized feedback.</li>\u0000 </ul>\u0000 <p>What this paper adds\u0000\u0000 </p><ul>\u0000 \u0000 <li>This study examines two types of GAI-assisted feedback (idea-oriented and task-oriented) and their impact on the dynamics and outcomes of online IBDs.</li>\u0000 \u0000 <li>Idea-oriented feedback significantly enhances cognitive presence and promotes deeper inquiry, while task-oriented feedback fosters social presence and group cohesion.</li>\u0000 \u0000 <li>GAI-assisted feedback, when aligned with specific learning objectives, can meaningfully promote IBD effective","PeriodicalId":48315,"journal":{"name":"British Journal of Educational Technology","volume":"56 5","pages":"1710-1734"},"PeriodicalIF":8.1,"publicationDate":"2025-05-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144809293","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}
<div> <section> <p>Virtual reality (VR) has been gaining prominence in education, with its interactive capabilities continually expanding. This quantitative study (<i>N =</i> 91) tested the educational effectiveness of high-immersion VR (HiVR) versus low-immersion VR (LiVR) and the impact of interactivity on vocabulary learning. The between-subjects portion of this study compared foreign language vocabulary learning using HiVR headsets and traditional laptops (LiVR). Multivariate analyses of covariance revealed that although the vocabulary scores of learners using HiVR were higher than the scores of learners using LiVR, the difference was not statistically significant. The within-subjects portion of this study tested the impact of the interaction with virtual objects representing the target vocabulary. Although students reported enjoying the interactive aspects of the experience, the interactivity did not significantly impact learning outcomes in either HiVR or LiVR. These findings have practical and theoretical implications about how different degrees of immersion and interactivity influence vocabulary learning and retention. The study is relevant for scholars and language teachers, as well as curriculum and VR application designers.</p> </section> <section> <div> <div> <h3>Practitioner notes</h3> <p>What is already known about this topic? </p><ul> <li>High-immersion virtual reality (HiVR) offers contextualized vocabulary learning through interacting with objects.</li> <li>Vocabulary is essential for language learning.</li> <li>Research on interaction with virtual objects has received little attention.</li> </ul> <p>What this paper adds? </p><ul> <li>It investigates the impact of VR on vocabulary learning.</li> <li>It explores the effect of object interactivity on vocabulary learning.</li> <li>It shows that VR improves vocabulary learning and retention regardless of object interaction.</li> </ul> <p>Implications for practice and/or policy </p><ul> <li>Our main implication is that VR can be beneficial for vocabulary retention.</li> <li>Students using HiVR and low-immersion VR (LiVR) make comparable learning gains.</li> <li>Practitioners should create VR activities that capitalize on the immersive features of the technology while keeping cognitive deman
{"title":"The impact of high-immersion virtual reality and interactivity on vocabulary learning","authors":"Regina Kaplan-Rakowski, Tricia Thrasher","doi":"10.1111/bjet.13603","DOIUrl":"https://doi.org/10.1111/bjet.13603","url":null,"abstract":"<div>\u0000 \u0000 <section>\u0000 \u0000 <p>Virtual reality (VR) has been gaining prominence in education, with its interactive capabilities continually expanding. This quantitative study (<i>N =</i> 91) tested the educational effectiveness of high-immersion VR (HiVR) versus low-immersion VR (LiVR) and the impact of interactivity on vocabulary learning. The between-subjects portion of this study compared foreign language vocabulary learning using HiVR headsets and traditional laptops (LiVR). Multivariate analyses of covariance revealed that although the vocabulary scores of learners using HiVR were higher than the scores of learners using LiVR, the difference was not statistically significant. The within-subjects portion of this study tested the impact of the interaction with virtual objects representing the target vocabulary. Although students reported enjoying the interactive aspects of the experience, the interactivity did not significantly impact learning outcomes in either HiVR or LiVR. These findings have practical and theoretical implications about how different degrees of immersion and interactivity influence vocabulary learning and retention. The study is relevant for scholars and language teachers, as well as curriculum and VR application designers.</p>\u0000 </section>\u0000 \u0000 <section>\u0000 \u0000 <div>\u0000 \u0000 <div>\u0000 \u0000 <h3>Practitioner notes</h3>\u0000 <p>What is already known about this topic?\u0000\u0000 </p><ul>\u0000 \u0000 <li>High-immersion virtual reality (HiVR) offers contextualized vocabulary learning through interacting with objects.</li>\u0000 \u0000 <li>Vocabulary is essential for language learning.</li>\u0000 \u0000 <li>Research on interaction with virtual objects has received little attention.</li>\u0000 </ul>\u0000 <p>What this paper adds?\u0000\u0000 </p><ul>\u0000 \u0000 <li>It investigates the impact of VR on vocabulary learning.</li>\u0000 \u0000 <li>It explores the effect of object interactivity on vocabulary learning.</li>\u0000 \u0000 <li>It shows that VR improves vocabulary learning and retention regardless of object interaction.</li>\u0000 </ul>\u0000 <p>Implications for practice and/or policy\u0000\u0000 </p><ul>\u0000 \u0000 <li>Our main implication is that VR can be beneficial for vocabulary retention.</li>\u0000 \u0000 <li>Students using HiVR and low-immersion VR (LiVR) make comparable learning gains.</li>\u0000 \u0000 <li>Practitioners should create VR activities that capitalize on the immersive features of the technology while keeping cognitive deman","PeriodicalId":48315,"journal":{"name":"British Journal of Educational Technology","volume":"56 6","pages":"2647-2670"},"PeriodicalIF":8.1,"publicationDate":"2025-05-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145248637","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}
Seyed Parsa Neshaei, Paola Mejia-Domenzain, Richard Lee Davis, Tanja Käser
Reflective writing is known as a useful method in learning sciences to improve the metacognitive skills of students. However, students struggle to structure their reflections properly, limiting the possible learning gains. Previous works in educational technologies literature have explored the paradigms of learning from worked and modelling examples, but (a) their application to the domain of reflective writing is rare, (b) such methods might not scale properly to large-scale classrooms, and (c) they do not necessarily take the learning needs of each student into account. In this work, we suggest two approaches of integrating AI-enabled support in digital systems designed around learning from worked and modelling examples paradigms, to provide personalized learning and feedback to students using large language models (LLMs). We evaluate Reflectium, our reflective writing assistant, show benefits of integrating AI support into the learning from examples modalities and compare the perception of the users and their interaction behaviour when using each version of our tool. Our work sheds light on the applicability of generative LLMs to different types of providing support using the learning from examples paradigm, in the domain of reflective writing.
Practitioner notes
What is already known about this topic
Reflective writing fosters metacognitive skills and improves learning gains and personal growth.
The learning from worked and modelling examples paradigms is effective for skill acquisition and applying the acquired knowledge.
Existing reflective writing assistants usually lack dynamic, AI-driven feedback or interactivity, limiting personalization and adaptability to each user's own needs in the learning process.
What this paper adds
It introduces Reflectium, an AI-enabled reflective writing assistant, integrating intelligent and interactive writing support for both the learning from worked and modelling examples paradigms.
It demonstrates the use of a fine-tuned large language model (LLM) for providing feedback in the learning from worked examples version, and an LLM-powered conversational agent simulating instructor interactions for the learning from modelling examples version.
It reports findings from a user study comparing the positive imp
{"title":"Metacognition meets AI: Empowering reflective writing with large language models","authors":"Seyed Parsa Neshaei, Paola Mejia-Domenzain, Richard Lee Davis, Tanja Käser","doi":"10.1111/bjet.13601","DOIUrl":"10.1111/bjet.13601","url":null,"abstract":"<div>\u0000 \u0000 \u0000 <section>\u0000 \u0000 <p>Reflective writing is known as a useful method in learning sciences to improve the metacognitive skills of students. However, students struggle to structure their reflections properly, limiting the possible learning gains. Previous works in educational technologies literature have explored the paradigms of learning from worked and modelling examples, but (a) their application to the domain of reflective writing is rare, (b) such methods might not scale properly to large-scale classrooms, and (c) they do not necessarily take the learning needs of each student into account. In this work, we suggest two approaches of integrating AI-enabled support in digital systems designed around learning from worked and modelling examples paradigms, to provide personalized learning and feedback to students using large language models (LLMs). We evaluate Reflectium, our reflective writing assistant, show benefits of integrating AI support into the learning from examples modalities and compare the perception of the users and their interaction behaviour when using each version of our tool. Our work sheds light on the applicability of generative LLMs to different types of providing support using the learning from examples paradigm, in the domain of reflective writing.</p>\u0000 </section>\u0000 \u0000 <section>\u0000 \u0000 <div>\u0000 \u0000 <div>\u0000 \u0000 <h3>Practitioner notes</h3>\u0000 <p>What is already known about this topic\u0000\u0000 </p><ul>\u0000 \u0000 <li>Reflective writing fosters metacognitive skills and improves learning gains and personal growth.</li>\u0000 \u0000 <li>The learning from <i>worked</i> and <i>modelling</i> examples paradigms is effective for skill acquisition and applying the acquired knowledge.</li>\u0000 \u0000 <li>Existing reflective writing assistants usually lack dynamic, AI-driven feedback or interactivity, limiting personalization and adaptability to each user's own needs in the learning process.</li>\u0000 </ul>\u0000 <p>What this paper adds\u0000\u0000 </p><ul>\u0000 \u0000 <li>It introduces Reflectium, an AI-enabled reflective writing assistant, integrating intelligent and interactive writing support for both the learning from <i>worked</i> and <i>modelling</i> examples paradigms.</li>\u0000 \u0000 <li>It demonstrates the use of a fine-tuned large language model (LLM) for providing feedback in the learning from <i>worked</i> examples version, and an LLM-powered conversational agent simulating instructor interactions for the learning from <i>modelling</i> examples version.</li>\u0000 \u0000 <li>It reports findings from a user study comparing the positive imp","PeriodicalId":48315,"journal":{"name":"British Journal of Educational Technology","volume":"56 5","pages":"1864-1896"},"PeriodicalIF":8.1,"publicationDate":"2025-05-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144811282","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}
扫码关注我们
求助内容：
应助结果提醒方式：
京公网安备 11010802042870号