Sources

17.10.2026. Die Coburger Studienmesse - eine Kooperationsveranstaltung der Hochschule Coburg und der Agentur für Arbeit Bamberg-Coburg - ist eine Karrieremesse für Abiturient*innen. Rund 60 Hochschulen und Unternehmen informieren hier über diverse Studiengänge und Duale Studienmöglichkeiten und stehen für einen ersten persönlichen Kontakt zur Verfügung. Eine Anmeldung ist nicht erforderlich. Veranstalter: Hochschule Coburg in Kooperation mit der Agentur für Arbeit Bamberg-Coburg. Link: https://www.hs-coburg.de/veranstaltung/coburger-studienmesse/ .

26.02.2027. Wer in die Region Berlin-Brandenburg studieren möchte, kann sich auf der Messe DeinStudium über Studiengänge informieren. Einen Tag lang stellen sich mehr als 35 öffentliche und private Hochschulen und Bildungseinrichtungen mit ihren Angeboten vor, außerdem sind Anbieter dualer Studiengänge vor Ort. Studienberater:innen informieren auf der Messe persönlich zu Studiengängen, Bewerbungen und Finanzierung. Fachkräfte geben Erfahrungen und Wissenswertes zu den Themen Studium allgemein, Hochschulen und Studiengängen in Vorträgen weiter. Außerdem gibt es praktische Einblicke in fachübergreifende Studienprojekte und einen Info-Point, wo sich Menschen mit Fluchtgeschichte über ihre Studienmöglichkeiten in Berlin und Brandenburg informieren können. Veranstalter: TUB/CO TU Berlin Corporate. Link: https://www.berlin.de/messen/10663897-7918641-deinstudium-in-berlinbrandenburg.html .

26.02.2027. Wer in die Region Berlin-Brandenburg studieren möchte, kann sich auf der Messe DeinStudium über Studiengänge informieren. Einen Tag lang stellen sich mehr als 35 öffentliche und private Hochschulen und Bildungseinrichtungen mit ihren Angeboten vor, außerdem sind Anbieter dualer Studiengänge vor Ort. Studienberater:innen informieren auf der Messe persönlich zu Studiengängen, Bewerbungen und Finanzierung. Fachkräfte geben Erfahrungen und Wissenswertes zu den Themen Studium allgemein, Hochschulen und Studiengängen in Vorträgen weiter. Außerdem gibt es praktische Einblicke in fachübergreifende Studienprojekte und einen Info-Point, wo sich Menschen mit Fluchtgeschichte über ihre Studienmöglichkeiten in Berlin und Brandenburg informieren können. Veranstalter: TUB/CO TU Berlin Corporate. Link: https://www.berlin.de/messen/10663897-7918641-deinstudium-in-berlinbrandenburg.html .

14.11.2026. Die Messe Stuzubi Hannover ist eine Karrieremesse für Abiturient*innen, Fachabiturient*innen und Realschüler*innen. Zahlreiche Unternehmen, Universitäten, Hochschulen und Akademien präsentieren sich auf der Stuzubi Messe Hannover den Schüler*innen mit ihren Ausbildungs- und Studienangeboten und stehen für einen ersten persönlichen Kontakt zur Verfügung. Neben regulären Studiengängen und Ausbildungsberufen wird auf der Studien- und Berufsmesse Stuzubi in Hannover ebenso das duale Studium vorgestellt, dass einen akademischen Abschluss mit einer betrieblichen Ausbildung verbindet. Veranstalter: Stuzubi. Link: https://stuzubi.de/messen/hannover/ .

31.10.2026. Die Messe Stuzubi Frankfurt am Main ist eine Karrieremesse für Abiturient*innen, Fachabiturient*innen und Realschüler*innen. Zahlreiche Unternehmen, Universitäten, Hochschulen und Akademien präsentieren sich auf der Stuzubi Messe Frankfurt am Main den Schüler*innen mit ihren Ausbildungs- und Studienangeboten und stehen für einen ersten persönlichen Kontakt zur Verfügung. Neben regulären Studiengängen und Ausbildungsberufen wird auf der Studien- und Berufsmesse Stuzubi in Frankfurt am Main ebenso das duale Studium vorgestellt, dass einen akademischen Abschluss mit einer betrieblichen Ausbildung verbindet. Veranstalter: Stuzubi. Link: https://stuzubi.de/messen/frankfurt/ .

17.10.2026. Die Messe Stuzubi Berlin Herbst ist eine Karrieremesse für Abiturient*innen, Fachabiturient*innen und Realschüler*innen. Zahlreiche Unternehmen, Universitäten, Hochschulen und Akademien präsentieren sich auf der Stuzubi Messe Berlin den Schüler*innen mit ihren Ausbildungs- und Studienangeboten und stehen für einen ersten persönlichen Kontakt zur Verfügung. Neben regulären Studiengängen und Ausbildungsberufen wird auf der Studien- und Berufsmesse Stuzubi in Berlin ebenso das duale Studium vorgestellt, dass einen akademischen Abschluss mit einer betrieblichen Ausbildung verbindet. Veranstalter: Stuzubi. Link: https://stuzubi.de/messen/berlin/ .

10.10.2026. Die Messe Stuzubi München Herbst ist eine Karrieremesse für Abiturient*innen, Fachabiturient*innen und Realschüler*innen. Zahlreiche Unternehmen, Universitäten, Hochschulen und Akademien präsentieren sich auf der Stuzubi Messe München den Schüler*innen mit ihren Ausbildungs- und Studienangeboten und stehen für einen ersten persönlichen Kontakt zur Verfügung. Neben regulären Studiengängen und Ausbildungsberufen wird auf der Studien- und Berufsmesse Stuzubi in München ebenso das duale Studium vorgestellt, dass einen akademischen Abschluss mit einer betrieblichen Ausbildung verbindet. Veranstalter: Stuzubi. Link: https://stuzubi.de/messen/muenchen/ .

10.10.2026. Die Messe Stuzubi Stuttgart Herbst ist eine Karrieremesse für Abiturient*innen, Fachabiturient*innen und Realschüler*innen. Zahlreiche Unternehmen, Universitäten, Hochschulen und Akademien präsentieren sich auf der Stuzubi Messe Stuttgart den Schüler*innen mit ihren Ausbildungs- und Studienangeboten und stehen für einen ersten persönlichen Kontakt zur Verfügung. Neben regulären Studiengängen und Ausbildungsberufen wird auf der Studien- und Berufsmesse Stuzubi in Stuttgart ebenso das duale Studium vorgestellt, dass einen akademischen Abschluss mit einer betrieblichen Ausbildung verbindet. Veranstalter: Stuzubi. Link: https://stuzubi.de/messen/stuttgart/ .

10.10.2026. Die Woche der Seelischen Gesundheit möchte auf die unterschiedlichen Strategien zur Bewältigung und auf das vielfältige psychosoziale Hilfsangebot in Deutschland aufmerksam machen sowie zum gemeinsamen Austausch und gegenseitiger Unterstützung aufrufen. Jährlich finden über 800 Events regional vor Ort und online statt. Ob Vorträge, Workshops, Podcasts oder Kunstausstellungen – alle sind eingeladen, bei verschiedenen Veranstaltungsformaten und Aktionstagen die kleinen und großen Angebote der psychiatrischen und psychosozialen Einrichtungen in ihrer Umgebung kennenzulernen. Die Veranstaltungen der Aktionswoche sind in der Regel kostenfrei und richten sich an alle Generationen. Sie tragen dazu bei, Berührungsängste abzubauen und Betroffene sowie deren Angehörige einzubinden. Die Angebote wollen jeden motivieren, sich Hilfe zu suchen und vor allem Präventions- und Beratungsangebote in der Nähe kennenzulernen. Veranstalter: Aktionsbündnis Seelische Gesundheit . Link: https://www.seelischegesundheit.net/aktionen/aktionswoche/ .

11.11.2026. Angesichts zunehmenden Drucks und Angriffen auf den demokratischen Rechtsstaat und einer immer diverseren Gesellschaft thematisiert der Fachtag aktuelle Herausforderungen zur Verflechtung von Demokratie, Religion und politscher Bildung. Im Zentrum der Diskussion stehen dabei zentrale Fragen des gesellschaftlichen Zusammenlebens, welche gemeinsam mit Expertinnen und Experten aus Wissenschaft und Praxis betrachtet werden. Während der Veranstaltung können sich Organisationen der religiös-informierten politischen Bildungsarbeit auf einem „Markt der Möglichkeiten“ präsentieren. Veranstalter: bpb. Link: https://www.bpb.de/veranstaltungen/veranstaltungskalender/581773/demokratie-und-religion-ressourcen-und-reibungsflaechen/ .

09.11.2026. In den Online-Vorträgen zum Orientierungsrahmen Globale Entwicklung zeigen die als Referierenden eingeladenen Autor*innen, wie Bildung für nachhaltige Entwicklung (BNE) in verschiedenen Fächern und in der Grundschule verankert werden kann. Dazu werden die Unterrichtsbeispiele vorgestellt und die fachdidaktischen Ansätze sowie Kompetenzen näher erläutert. Ergänzend haben Sie Gelegenheit Ihre individuellen Fragen zu stellen. Veranstalter: Engagement Global gGmbH. Link: https://www.engagement-global.de/de/veranstaltung-detail/14716--bne-mit-globaler-perspektive-in-der-grundschule .

26.10.2026. In dieser digitalen Hospitation werfen die Teilnehmenden gemeinsam einen Blick in die KITA21-Kita "Kleine Wunder" in Schönkirchen – eine Einrichtung, die Bildung für nachhaltige Entwicklung (BNE) aktiv im Alltag lebt. Im Mittelpunkt stehen echte Erfahrungen aus der Praxis: Was funktioniert? Was hat die Kinder begeistert? Und wie lässt sich das auf die eigene Einrichtung übertragen? Veranstalter: S.O.F. Save Our Future – Umweltstiftung. Link: https://www.kita21.de/mitmachen/kita21-in-sh/weitere-angebote/ .

06.11.2026. Wie beeinflusst die Klimakrise das Leben – heute und in Zukunft? Wie wird aus Wissen gemeinsames Handeln? Gemeinsam werden Klimabildungsangebote weiterentwickelt und Möglichkeiten diskutiert, die möglichst vielen Menschen zugänglich gemacht werden. Das kostenfreie Netzwerktreffen bietet Gelegenheit, aktuelle Projekte und Methoden kennenzulernen, neue Impulse zu gewinnen sowie sich mit anderen Akteurinnen und Akteuren der Klimabildung auszutauschen und zu vernetzen. Veranstalter: ANU Bundesverband e. V.. Link: https://www.umweltbildung.de/arbeitskreis-klimabildung/details/netzwerktreffen-im-arbeitskreis-klimabildung .

21.10.2026. Dieses Jahr wird für angehenden pädagogischen Fachkräften und ihren Lehrkräften oder Ausbildungsanleitenden sowie allen interessierte Fachkräften aus Hamburger Kitas ein besonderes Fortbildungsformat angeboten: In der Fortbildung wird sich dem Thema "Die Kita als Lernort für Natur- & Ressourcenschutz" mit vielseitigen interaktiven Methoden theoretisch an. In der digitalen Hospitation geben Kita-Leitungen Einblick in Ihre Arbeit und stehen für Fragen zur Verfügung. Veranstalter: KLIMAfuchs. Link: https://klimafuchs-kita.de/termine/fortbildung-auszubildende/ .

15.10.2026. In diesem Online-Workshop wird eine Methodenauswahl für die Jugend- und Erwachsenenbildung zum Thema Klimagerechtigkeit vermittelt. Es werden niedrigschwellige Methoden rund um das Thema Klimagerechtigkeit erprobt. Es handelt sich hierbei um interaktive Bildungsmethoden Veranstalter: AWO International e. V.. Link: https://www.globaleslernen.de/de/veranstaltungen/methodenworkshop-klimagerechtigkeit-0 .

15.10.2026. Die aktuellen Kriege betreffen junge Menschen in besonderer Weise: Sie sollen Wehrdienst leisten. Die Schulden für Aufrüstung gehen zulasten der Investitionen in ihre Zukunft, zum Beispiel in Bildung und Klimaschutz. Und auch im Schulalltag spiegeln sich internationale Konflikte wie in der Ukraine oder in Nahost längst wider. Frieden ist längst kein abstraktes Thema mehr in entfernten Regionen oder Zeiten, sondern ein zentrales, hochumstrittenes Thema. Wie kann Bildungsarbeit auf diese Themen eingehen? Welche Ansätze und Angebote gibt es? Veranstalter: Entwicklungspolitisches Netzwerk Hessen e. V. (EPN Hessen) in Kooperation mit Pro Peace – Forum Ziviler Friedensdienst e. V.. Link: https://eveeno.com/druebergebabbelt .

12.10.2026. Geschichten können Kinder darin bestärken, sie selbst zu sein, ihnen Mut machen und sie motivieren, aktiv gegen Ungerechtigkeit zu werden. Aber wo finde ich solche Geschichten und wie kann ich sie auswählen? In der Fortbildung lernen die Teilnehmenden, wie Bücher und andere Medien dazu beitragen, dass Kinder Bilder von sich selbst und der Welt entwickeln und durch Bücher bestärkt werden können. Veranstalter: AWO International e. V.. Link: https://www.awointernational.de/mitmachen/globales-lernen/fortbildungen .

11.10.2026. Rassismus ist in den letzten Jahren immer salonfähiger geworden: AfD im Bundestag, Einwanderungsdebatte und vieles mehr verschärft eine offene Polarisierung in "Wir" und "Die". Sprechen Betroffene oder Beobachter*innen Rassismus an, wird jedoch oft geantwortet, man habe es doch gar nicht so gemeint. Wie können wir uns dazu verhalten? Muss wirklich auf alles reagiert werden? Wo fängt Rassismus eigentlich an? Und was bedeutet das für unseren beruflichen und privaten Alltag? Diesen und weiteren Fragen geht das Seminar nach. Veranstalter: DGB Bildungswerk e. V.. Link: https://www.dgb-bildungswerk.de/bildungsurlaub/seminare/mit-rassismus-im-alltag-souveraen-umgehen .

07.10.2026. Mit Bildung für nachhaltige Entwicklung (BNE) lernen wir, die Zukunft nachhaltig zu gestalten. Das Landesprogramm Schule der Zukunft unterstützt Schulen dabei, BNE dauerhaft in Unterricht und Schulentwicklung zu verankern. Diese Informationsveranstaltung richtet sich an neu teilnehmende und interessierte Schulen. Die Teilnahme ist kostenfrei. Veranstalter: Abenteuer Lernen e. V.. Link: https://www.abenteuerlernen.org/angebote/fortbildungen/bne-fortbildungen-test/landesprogramm-schule-der-zukunft-online-informationsveranstaltung/individual-registration .

27.11.2026. Wie können digitale Lösungen den Schulalltag konkret erleichtern? Micha Pallesche (Rektor Ernst-Reuter-Gesamtschule) gibt Einblicke in aktuelle Projekte aus der Schulpraxis und zeigt anhand konkreter Beispiele, wie digitale Anwendungen Wissen zugänglicher machen, Abläufe vereinfachen und Kollegien entlasten können. Dabei geht es auch um neue Möglichkeiten durch Künstliche Intelligenz: von selbst entwickelten Anwendungen ohne klassische Programmierkenntnisse über datenschutzsensible Lösungen bis hin zur Frage, wie vorhandene Daten sinnvoll für die Schulentwicklung genutzt werden können. Im Mittelpunkt stehen praxiserprobte Ideen, die sich auf andere Schulen übertragen und weiterdenken lassen. Veranstalter: Bitkom e.V.. Link: https://www.bitkom.org/Fortbildungsreihe-Digitale-Resilienz-im-Schulalltag/4-Phase .

This study examined the impact of learning analytics-based personalized feedback on student engagement behavior, learning outcomes, and perceptions of the feedback in a large introductory STEM course. Students were randomly assigned to either an experimental group, which received four personalized feedback messages between Weeks 6 and 14, or a control group, which received generic messages consistent with the instructor’s previous teaching practices. The feedback was personalized based on students’ engagement with homework and practice problems and was designed to promote specific behaviors in these two areas. Results showed that the experimental group exhibited greater engagement in the targeted behaviors compared to the control group. However, no significant differences were found between the groups in perceived helpfulness of the feedback or learning outcomes. This study addresses the limitations of the current implementation of personalized feedback and highlights the value of integrating behavioral data with learning outcomes and self-report responses to more accurately evaluate its impact.

Reading comprehension remains challenging in secondary language education, particularly in digital environments that provide limited adaptivity, interactivity, and multimodal support. This study developed and evaluated a multimodal e-learning platform integrating artificial intelligence (AI), game-based learning (GBL), and augmented reality (AR). The AI module provided adaptive feedback and vocabulary scaffolding, the GBL component offered interactive comprehension tasks, and the AR feature visualized textual content. A quasi-experimental mixed-methods design involved 214 tenth-grade language learners allocated to an experimental group (n = 107) or a control group (n = 107). During the six-week intervention, the experimental group used the platform for reading instruction, whereas the control group received conventional teaching. Data were collected through a pilot-tested 25-item reading comprehension test, a student engagement and motivation scale, and semi-structured interviews. The experimental group’s mean comprehension score increased from 78.6 to 90.2, compared with an increase from 78.3 to 82.1 in the control group. Analysis of covariance indicated a significant group effect, F(1, 211) = 66.08, p < .001, partial η2 = .24. Hierarchical regression showed that platform use was significantly associated with comprehension gain (final adjusted R2 = .45), and exploratory component-level associations were strongest for AI-use frequency (β = 0.42, p < .001). Interview findings indicated greater motivation, confidence, and cognitive engagement, alongside reduced reading anxiety. These findings suggest that coordinated AI, GBL, and AR features can support cognitive and motivational dimensions of secondary reading instruction.

This study considers the professional needs of instructional design (ID) and educational technology (ET) professionals, including not only the technical and domain-related knowledge and skills traditionally covered by most professional organizations and institutions, but also crucial professional and cross-disciplinary skills and dispositions, with a surprisingly large emphasis on the latter two themes. ID and ET program graduates seek to be competent to stay competitive in the workplace. However, the constantly expanding and evolving field and changing job expectations present a challenge. This leads to an important question of whether what current and recent graduates have learned in their program is sufficient to be considered competent and up-to-date and what other areas of need may be present in the curriculum. The study’s findings are based on the results of 31 interviews with ID and ET professionals from the three contexts (K12, higher education, and corporate environments). Data were analyzed using a collaborative Constant Comparative Method for Naturalistic Inquiry, involving iterative coding and thematic refinement to identify key professional competencies and dispositions. While the knowledge and skills identified aligned well with the current literature in the area, the cross-disciplinary and dispositions sections revealed findings that will help program directors and designers, faculty, employers, and current and future graduates.

The integration of generative AI (GAI) into collaborative learning has drawn increasing attention for its potential to reshape group dynamics and support collaborative knowledge construction (CKC). Yet, how CKC unfolds in GAI-supported contexts, particularly regarding interaction patterns, regulatory processes, and task outcomes, remains insufficiently understood. Grounded in group-regulated learning perspectives, this study examines how GAI -supported collaborative processes relate to students’ collaborative knowledge construction. The study analyzes detailed interaction data collected from three-member collaborative groups across two face-to-face tasks, yielding 24 group-task instances. Using a multi-method analytical framework that combines discourse analysis, hierarchical clustering, epistemic network analysis, process mining, and summative assessments, we identify four exploratory collaborative patterns: (1) the positive regulation-oriented pattern (Cluster 1), characterized by medium-level performance and more complex knowledge construction; (2) the negative regulation-oriented pattern (Cluster 4), also linked to medium-level performance and more complex CKC; (3) the socio-emotionally driven regulation pattern (Cluster 2), associated with low-level performance and simpler knowledge construction; and (4) the strategy-with-AI and monitoring-oriented regulation pattern (Cluster 3), linked to high performance and less complex CKC than Clusters 1 and 4, yet still showing more complexity than Cluster 2. Based on these findings, the study concludes with pedagogical implications, providing guidance for educators and instructional designers to optimize GAI-supported collaborative learning, fostering greater student engagement, and enhancing knowledge construction.

This study investigated recall performance in contextual and context-free tasks designed within an immersive virtual reality (IVR) learning environment in relation to visual-spatial memory (VSM) capacity. Using a 2 × 2 mixed factorial design, 53 university students completed both contextual and context-free recall tasks and were categorized as having high or low VSM capacity. The results revealed that recall performance was significantly higher in contextual tasks compared to context-free ones. However, VSM capacity did not have a significant main effect on recall performance, nor was there a significant interaction between task type and VSM capacity. Gender was examined as an exploratory variable and showed no significant association with VSM capacity. These findings suggest that contextual richness in IVR environments can facilitate recall across learners with different levels of VSM capacity. Overall, the study highlights the pedagogical potential of IVR environments and underscores the importance of contextual instructional design in supporting recall for spatially complex content.

Educational innovations should articulate a clear rationale that situates the innovation within existing literature. Current frameworks, such as ESSA Tier IV, emphasize this requirement, though they only mandate a general review of research. Consequently, the process for demonstrating this connection currently lacks standardization, leading to variability in the quality and depth of the rationale provided across different educational tools and interventions. The aim of this paper is to introduce a structured methodological approach for evaluating educational innovations in relation to their supporting literature, emphasizing a systematic process that balances conceptual rigor with methodological plurality. We present the development and refinement of our methodological approach and illustrate its application through the method of two case studies. Our findings are structured around five systematic steps, integrating rapid evidence reviews, a weighted assessment of findings, and alignment with established principles from the Science of Learning. We conclude that such an approach can form part of pathways to foster more effective collaboration between researchers and developers, thereby strengthening the design and impact assessments of educational innovations.

Multimodal Learning Analytics (MMLA) enables the collection of extensive data on students’ learning; however, its ecological validity in authentic classroom environments remains limited. Systems validated in controlled laboratory settings frequently encounter challenges when applied in complex, dynamic classrooms. To address this gap, this study presents a field-based evaluation of a classroom-native Active Learning Classroom (ALC) system designed to capture and integrate behavioral, cognitive, and affective data streams. These data were used to derive instructionally relevant indicators of engagement, attention, and satisfaction. A quasi-experimental design was employed to investigate two primary questions: (1) whether the ALC-derived indicators form a valid and interpretable measurement structure, and (2) whether learning outcomes differ between the ALC cohort (N = 112) and a historical control cohort (N = 121). Confirmatory factor analysis supported the proposed student-level measurement structure. Team–unit level analysis revealed significant small-to-moderate relationships between behavioral and cognitive indicators, indicating that greater observable participation does not necessarily reflect deeper cognitive elaboration. Learner-generated state signals from the real-time Response-to-Instruction (RTI) system corresponded with variations in behavioral, attentional, and discourse-based cognitive indicators, providing contextual anchors for interpreting automated multimodal analytics. The ALC cohort demonstrated significantly higher unit-level posttest performance compared to the historical control cohort. Overall, these findings provide preliminary support for the instructional design and implementation of classroom-native MMLA, demonstrating how ecologically grounded analytics can transform multimodal classroom data into interpretable, theory-informed insights about learning processes in authentic higher classroom settings.

Digital tools are widely used in today’s classrooms to enhance student learning. While many of these tools are designed with epistemic underpinnings, their epistemic features and roles may not be obvious to or taken up by learners, which could affect students’ inquiry and learning using the tools. This study examined whether providing epistemic scaffolding to prompt students to reflect on the tool’s epistemic features and roles could improve their learning, and how students’ epistemic reflection related to their conceptual understanding and inquiry practices while using the tool. Four 8th-grade classes used a digital text tool (i.e., VidyaMap) to study photosynthesis and energy transformation. Two classes (n = 49) received epistemic scaffolds (i.e., epistemic reflection prompts), while the others (n = 51) served as a comparison group without epistemic scaffolds. Results showed that students who received epistemic scaffolding demonstrated greater gains in conceptual understanding than those in the comparison group. Within the experimental group, students’ epistemic reflection scores predicted their post-conceptual understanding after controlling for pre-conceptual understanding. Qualitative analyses further showed that students with higher levels of epistemic reflection engaged in more constructive inquiry practices while using the tool than those with lower levels of epistemic reflection, providing process-level evidence of a possible mechanism through which epistemic reflection may support learning. The findings highlight the importance of making the epistemic features and roles of digital tools explicit in technology-supported learning environments.

This study developed and validated a suite of Generative Artificial Intelligence Readiness and Perception (GenAI-RP) Scale for faculty and students in higher education. Three subscales, including a) Readiness, b) Benefit, and c) Challenge were created based on a review of relevant literature and validated through content experts’ review and confirmatory factor analysis (CFA). The Readiness scale includes three factors: a) GenAI Comprehension, b) Ethical Awareness of GenAI, and c) GenAI Utilization and Proficiency. The Benefit scale consists of two factors: a) Effectiveness and b) Empowerment. Lastly, the Challenge scale has three factors: a) Ethics and Privacy Concerns, b) Negative Educational Impact, and c) Accuracy and Sensitivity. All scales demonstrated a satisfactory model fit for both groups based on CFA, except the challenge scale for faculty.

While prior research has demonstrated the motivational affordances of AI-assisted storytelling, limited empirical work has examined how GenAI can be systematically aligned with a coherent learning theory to scaffold literacy development across cognitive and affective dimensions. Addressing the lack of theory-aligned AI literacy designs, the intervention operationalized Interest-Driven Creator theory loop via three GenAI roles: a dialogic reading companion, a multimodal co-creator generating illustrations from students’ retellings, and an evaluator providing criterion-referenced feedback on five narrative components. Across an eight-week intervention, pre–post assessments showed significant gains in overall narrative proficiency, particularly in structural dimensions (introduction and conclusion), while improvements in mental-state representation and overall message were positive but nonsignificant. Affective outcomes indicated a significant increase in triggered situational interest and stable flow experiences. Action-log and profile analyses further showed that students engaged with the system through iterative image generation, repeated evaluations, and peer viewing. Correlational analyses revealed that creation-phase behaviors were more associated with narrative gains, interest, and flow than learning or sharing activities. These findings illustrate how theory-aligned GenAI may integrate multimodal creation and formative feedback to support creation-centered literacy development.

Although research indicates that Digital Personalised Learning (DPL) can improve aspects of early-years literacy and numeracy in low- and middle-income countries (LMICs), it is necessary to determine whether teacher-facing learning analytics dashboards further impact education outcomes. Specifically, there is a need to evaluate different dashboard designs to determine which design features best support pedagogical decision-making, given that evidence on how such dashboards are used in LMICs remains limited. This study is the first to explore how different dashboards, designed to present teachers with learners’ foundational literacy and numeracy competency levels, can impact device usage, learning outcomes and teachers’ evaluation of learner progress. An A/B/C test in 4487 Kenyan schools compared three randomly assigned school partitions: no dashboard, an algorithm-generated List Dashboard and an algorithm-generated Grouped Dashboard. Impact was evaluated based on learners’ usage of DPL and its impact on literacy and numeracy outcomes. The alignment between algorithm-generated rankings and teachers’ ratings of these rankings was also investigated to determine how practitioners make sense of different data presentations. Findings demonstrate that learners in the Grouped Dashboard partition used the DPL tool more than those in the other two partitions. They also scored statistically significantly higher on two literacy summative test units, but not on numeracy test units. However, despite its potential benefits, the Grouped Dashboard was perceived as being less accurate in determining learner level according to teacher ratings. This research is significant in bridging the gap between teacher-facing dashboard design and its intended impact on teacher sensemaking, learner device usage and learning outcomes in DPL environments in LMICs. Outcomes rigorously inform future design decisions on teacher-AI collaboration and highlight the potential of A/B/n testing to consider cultural and regional factors in technology-supported instruction.

Supporting student engagement is crucial, as it is a key predictor of academic success in online learning. However, little is known about whether interventions can disrupt self-reinforcing engagement patterns such as the Matthew effect. This study investigates whether personalised feedback on attendance can increase engagement levels and reduce engagement disparities in an online learning environment. We conducted an experiment in which one group of students received personalised reports with attendance analytics, a list of missed webinars for asynchronous viewing, and recommendations to attend live sessions, while a control group received no such support. Results confirmed that prior engagement strongly predicted subsequent participation, consistent with a Matthew effect pattern. The support was associated with modest improvements in synchronous engagement and more stable high participation, while showing no effect on asynchronous engagement. The intervention did not decrease persistence of low engagement, which indicates that engagement gaps persisted. These findings suggest that while personalised support may help sustain synchronous and high engagement to a certain degree, it is insufficient to disrupt cumulative disadvantages in participation.

The rapid expansion of artificial intelligence (AI)–based writing tools in education has outpaced systematic understanding of how pedagogical intentions, theoretical commitments, and technical customization are translated into AI system design. Customization is conceptualized here as the deliberate modification of AI system behavior to align outputs with pedagogical intent. In response to this gap, this systematic review synthesizes 23 empirical studies from recent AI-in-education research to examine how AI systems are customized for writing pedagogy. Specifically, the review analyzes patterns of alignment among pedagogical goals, theoretical principles, customization methods, and implementation modes. Findings indicate that although pedagogical aims have increasingly shifted beyond surface-level correction toward writing processes, feedback literacy, and higher-order development, customization practices remain predominantly driven by performance-oriented theories and technical optimization. Prompt engineering and fine-tuning dominate existing approaches, with limited evidence that learning-centered, sociocultural, or developmental theories are operationalized within AI system behavior. Across studies, customized AI systems demonstrate benefits in improving writing quality, learner engagement, and assessment reliability; however, persistent challenges related to feedback inaccuracy, learner overreliance, and ethical transparency are consistently reported. By identifying systematic misalignments and underexplored research spaces, particularly in longitudinal design, interactional analysis, and theory-informed customization, this review reveals a structural gap between educational aspirations and current AI design practices, specifically the disconnection between learning-oriented pedagogical goals and the performance-driven theories and modular customization techniques that predominantly operationalize them. The findings underscore the need to move beyond automation-focused models toward developmentally oriented, theoretically grounded AI systems capable of supporting sustained learning in writing pedagogy.

Systematic reviews are time-consuming endeavors that require knowledgeable human reviewers to screen studies for relevance and extract data following a specific coding scheme before any analysis or synthesis can occur. Large language models (LLMs) hold promise for substantially accelerating this process and reducing reviewer workload, yet their application within the context of systematic reviews in the field of education remains underexplored. We address this issue in two ways: through empirical studies and the iterative development of an open-source software tool. First, we conducted two empirical studies examining the efficacy of using LLMs for data extraction using data from a published review on pedagogical agents. We extracted a variety of data types from 112 studies and compared the results to data extracted by human coding. Results indicate that LLMs struggled with extracting data accurately and therefore are not ready to be used as primary data extraction tools without explicit human validation of the data extracted. These findings highlight the dire need for a human-in-the-loop (HIL) approach to AI-assisted data extraction. We then propose a HIL workflow and introduce and describe the development of a free, web-based, open-source tool designed to support user-friendly, human-validated data extraction with LLMs.

This study evaluates the implementation of Augmented Reality (AR) and Virtual Reality (VR) applications for teaching electronic components to vocational high school students in Bandung, Indonesia. The study aims to determine whether these technologies can boost students’ interest, interaction, and engagement. Fifty Grade-XI electronics students (47 male, 3 female) participated in a single-group classroom session using the AR/VR application and then completed questionnaires and follow-up interviews. The research employed a Usability Evaluation approach to assess the applications’ ease of use and effectiveness in learning electronic components. Data collection methods included questionnaires to gather quantitative insights into usability, perceived usefulness, student attitudes, intentions, and perceived ease of use. Interviews are conducted to explore students’ experiences in depth. Questionnaire ratings on a five-point Likert scale indicated positive perceptions of usefulness and ease of use. The interview findings suggested higher engagement and easier visualisation of abstract circuit concepts. However, two practical challenges are noted: intermittent internet connectivity and limitations of lower-spec devices. Overall, the findings suggest that a well-designed AR/VR tool with strong usability can enhance student motivation and conceptual understanding in vocational electronics classes, provided that connectivity and device constraints are addressed effectively.

Recognizing Chinese characters is a major challenge for beginning second-language learners, particularly those from alphabetic-language backgrounds, because its logographic writing system lacks grapheme–phoneme correspondence. Although intelligent tutoring systems (ITSs) have proven effective in domains such as mathematics and physics, their application to Chinese character learning remains underexplored. The purpose of this research was to 1) explore the design and usability of an interface of an intelligent tutoring system for recognizing Chinese characters, and 2) compare the pedagogical effectiveness of two forms of information presentation and feedback. A prototype system (an iPad Chinese character tutor) was developed and evaluated. Two groups were given 34 Chinese characters and phrases to learn using two different versions of the system. Version A contained a metaphor-based pedagogy, feedback, and extra instructions; Version B offered a simplified interface that focused on repetition. Participants’ learning performance and survey results were used to measure the effectiveness and usability of the system. Learning performance of the group that used Version A was statistically significantly higher than that of the Version B group. Participants rated Version A significantly higher than Version B on usability, satisfaction, functionality, and usefulness. This study lays the foundation for the development of an intelligent tutoring system (ITS) for Chinese character learning and highlights the value of integrating meaning-based, visual, and feedback-rich instruction.

The potential benefits of tracing on a computer screen with varying tactile experiences are unknown. To address these issues, an experiment was conducted to compare stylus tracing on a computer screen and finger tracing on a computer screen using a one-way, between-subjects design. A total of 75 college students were recruited and randomly assigned in equal numbers to one of the conditions: stylus tracing, finger tracing, and no tracing. The results showed that, compared to no tracing, stylus tracing and finger tracing improved learners’ comprehension and transfer. However, these learning benefits are also accompanied by heightened intrinsic cognitive load. Moreover, there were no significant differences between stylus tracing and finger tracing in terms of learning outcomes, cognitive load, or intrinsic motivation. These findings, along with educational implications, limitations, and future directions, are discussed within the frameworks of cognitive load theory and engagement theory.

This study examines how language teachers in selected urban Chinese contexts interpret and respond to generative artificial intelligence (GenAI) in their teaching. Drawing on interviews, focus groups, instructional artefacts, reflective memos, and teaching activity logs from 35 English and Japanese teachers, the study explores why teachers at different career stages adopt, limit, or avoid GenAI in language instruction. The findings show that early-career teachers were more likely to use GenAI pragmatically for lesson preparation, differentiated materials, and feedback support, whereas experienced teachers more often raised concerns about student authorship, effort, assessment fairness, and professional responsibility. These differences were not explained by technical competence alone. Teachers’ decisions depended on whether GenAI was perceived as pedagogically appropriate, institutionally supported, and ethically acceptable. Using the Pedagogical Activation through Interpretive Thresholds (PAIT) framework developed for this study, the analysis explains GenAI integration as a process of interpretive alignment rather than simple adoption. The study contributes to research on AI in language education by showing how teachers’ GenAI practices are mediated by disciplinary values, institutional ambiguity, and generational sensemaking.

Generative AI pedagogical agents can provide adaptive instructional dialogue, but their responses may also appear impersonal, overly formal, or insufficiently engaging. One possible design response is to embed instructionally relevant humor into agent–learner interaction. This study examined whether a humorous social-affective instructional style that combined content-related, non-disparaging humor with mild colloquial expression and encouraging language influenced learners’ self-reported cognitive load, immediate post-test performance, and perceived agent value in an online learning activity. A randomized experiment was conducted with 120 Chinese university students. Instructional style was experimentally manipulated, and learner gender was examined as a measured moderator. Compared with learners who interacted with the neutral-serious agent, learners who interacted with the humorous social-affective agent reported lower perceived intrinsic cognitive load, evaluated the agent more favorably, and performed better on an immediate post-test. An Instructional Style × Gender interaction was observed for perceived intrinsic cognitive load: the reduction was more evident among female learners than among male learners. Male learners, however, reported higher perceived agent value overall. These findings suggest that a carefully constrained humorous social-affective instructional style may be a useful but context-dependent design approach for generative AI pedagogical agents. The results should be interpreted cautiously because prior knowledge was not directly measured, cognitive load was assessed through self-report, and the study used one learning topic and one AI platform. The study contributes to research on GenAI-supported learning by clarifying how a humorous social-affective instructional style can be operationalized and evaluated as part of instructional agent design.

Integrating technology into higher education teaching is reshaping pedagogical practices and prompting academics to (re)construct their professional identities. This qualitative case study explores how academics at a South African university respond to these changes, drawing on identity theory and the Technological Pedagogical Content Knowledge (TPACK) model to analyse the data. Data were collected from 12 purposively selected academics through semi-structured interviews and were analysed thematically. Findings indicate that academics are compelled to transition from traditional to interactive, student-centred teaching, requiring ongoing adaptation to emerging digital tools, multimedia integration, and data analytics for student monitoring. However, this shift also introduced identity tensions driven by increased workload, student disengagement, and recurring technical issues. Despite these challenges, the participants reported enhanced pedagogical effectiveness and renewed engagement through digital methods. The study highlights the dual nature of technology integration as both a disruptor and an enabler of professional growth and argues for comprehensive institutional support structures, including professional development and technical assistance. The findings contribute to the discourse on teacher identity in digitally evolving higher education environments and offer practical implications for policies and professional learning.

Mastering skills in designing valid, controlled experiments is foundational for scientific reasoning and inquiry. Whereas prevailing interventions to develop experimentation skills often induce cognitive conflict through comparing others’ invalid experiments with valid ones, this study tested the hypothesis that deliberately committing and correcting one’s own errors in invalid designs yields better learning and transfer than studying others’ errors. University students (N = 276) were instructed on experimental design principles (e.g., control of variables), then practiced applying them using one of three learning methods: derr-explain-fix (intentionally designing an invalid experiment then explaining and correcting one’s deliberate errors) or spot-explain-fix (spotting, explaining, and correcting others’ deliberate errors in invalid designs) or errorless practice (designing a valid experiment and explaining it). All students were then tested on their near and far transfer in designing an experiment that was structurally analogous to or more complex than the one during practice. They were also assessed on critiquing a given design’s validity, and their negative knowledge in anticipating problem-specific design errors. Both errorful learning methods outperformed errorless practice on the near transfer, design critique, and error anticipation tests. Crucially, the derr-explain-fix method improved near and far transfer more than the spot-explain-fix method, which did not yield better far transfer than errorless practice. Yet, students lacked metacognitive awareness of the benefits of deliberate erring and, more broadly, errorful learning. Overall, these findings demonstrate that intentionally designing and correcting one’s own invalid experiments during practice is an effective way to enhance learning and transfer of experimental design principles.

Rubrics are widely used in education to clarify expectations and guide student performance. However, the optimal timing of rubric presentation — whether before, after, or at both stages of a task — remains underexplored. This pre-registered experimental study aimed to determine whether rubrics function more effectively as pre-task guides or post-task feedback tools by examining differences in writing performance and cognitive load. Additionally, it explored potential differences between sixth and eleventh graders. Participants included 271 students from a private school in the Northeastern United States, with 141 eleventh graders and 130 sixth graders. Students wrote persuasive essays under one of four conditions: rubric before, rubric after, rubric before and after, or no rubric. Performance was assessed at two time points: the initial draft (T1) and the revised draft (T2). Cognitive load (perceived difficulty and effort) was also measured. Results revealed that rubrics provided before did not have any effect on the initial draft quality (T1), while rubrics provided after or at both stages enhanced revision outcomes (T2). Eleventh graders showed greater improvement than sixth graders. Cognitive load varied across conditions, with earlier rubric access reducing perceived difficulty. Thus, the timing of rubric presentation significantly influences writing performance and cognitive engagement. Rubrics are most effective when provided after the task, serving as an effective post-task feedback tool.

In recent years, the importance of developing instructional designs in schools to enhance students’ life skills has received increased attention. Placing life skills related issues on the curriculum implies that students’ personal lives and experiences are being given a more prominent role in education. However, while existing empirical studies have revealed valuable insights into aspects relevant to the practice of teaching life skills, there is lack of research on how teachers guide students in situ within life skills education settings. In this study, we explore how teachers orient to students’ personal experiences, ideas and assumptions from their everyday lives during classroom conversations in the context of life skills education. Specifically, we examine the opportunities and challenges teachers might face in this type of instructional work. Our analysis draws on video data from a life skills project in which lower secondary school students and their teachers engaged in classroom conversations about issues related to youth, identity and belonging. The analysis shows (1) how the teachers sometimes used students’ personal experiences as resources for guiding their conceptual understanding, (2) the efforts required of teachers to handle emotional and sometimes sensitive student contributions in whole-class conversations and (3) the challenges teachers sometimes faced in integrating conceptual knowledge with students’ everyday experiences. We discuss these findings in relation to the existing literature.

The present study investigated the effects of instructor presence in a multimedia learning situation where seeing the instructor was deemed relevant to the learning content and requirements, in this case a language learning setting. Fifty Japanese university students were asked to learn from two videos on the four Mandarin tones, one with the instructor video present and one without. While viewing the videos, participants’ eye movements and skin conductance level were recorded, and measurements of learning performance and perceptions about learning were taken afterwards. Results showed no significant difference in overall learning performance when comparing the conditions with and without instructor presence, but learners indicated that they were more satisfied and perceived more learning when the instructor video was included. This suggests that instructor presence can influence learners’ attention and perceived learning. However, in the present study, videos of instructors modeling tone production without providing other cues to tone production did not lead to measurable improvements in learning outcomes, even though learners attended to the instructor videos and perceived them as helpful. These findings indicate that instructor presence is unlikely to improve learning outcomes unless the instructor’s visible behavior provides information learners can directly use for the task at hand. More broadly, the benefits of instructor presence may depend not only on the alignment between the information conveyed and the requirements of the learning task, but also on learners’ ability to recognize and use that information effectively.

Socially shared regulation of learning (SSRL) can promote high-level collaborative learning. However, the sequential interplay between sociocognitive, socioemotional and motivational group-level challenges and SSRL strategies remains unclear, particularly in comparative analyses of pre-service and in-service teachers engaged in computer-supported collaborative problem-solving tasks. To address this gap, 33 pre-service and 33 in-service teachers were video-recorded in face-to-face settings while working in groups of three on collaborative problem-solving tasks within a virtual simulation on a desktop computer. A total of 25.7 h of video data were analysed using a theory-driven coding scheme that captured sociocognitive, socioemotional and motivational challenges, as well as collaborative regulatory activities grounded in the SSRL framework. Process mining was employed to visualise and compare sequential patterns of challenges and regulation. The findings indicate that pre-service teachers exhibited stronger sequential patterns linking planning, monitoring and controlling activities, whereas in-service teachers demonstrated more pronounced sequences involving monitoring and planning following socioemotional and motivational challenges. These differences reflect distinct tendencies in how collaborative learning activities were organised over time, rather than differences in effectiveness or learning outcomes. This study contributes to the learning sciences by extending SSRL research and offering theoretical and methodological insights into analysing collaborative learning as a dynamic process. From a practical perspective, the findings inform the pedagogical design of collaborative learning and teachers’ professional development by highlighting the need for profile-sensitive scaffolds that support shared planning, monitoring and adaptive regulation.

Teacher academic and emotional support play a crucial role in students’ reading achievement, yet the underlying mechanisms remain underexplored. Utilizing data from PISA 2018, this study investigates how school belonging moderates the indirect effects of teacher academic and emotional support on reading achievement via mastery goal orientation. Based on a sample of mainland Chinese students (N = 12,058), the findings indicate that: (a) both academic and emotional teacher support positively predict reading achievement; (b) mastery goal orientation mediates these relationships; and (c) contrary to expectations, school belonging negatively moderates the indirect effects – the indirect effects are stronger when school belonging is low. For academic support, the first‑stage interaction (support → mastery) was significantly negative, whereas for emotional support, neither path interaction reached significance; nevertheless, the overall indices of moderated mediation were significant for both models, indicating that the joint contribution of the two stages (weighted by the strong main effects) produces a reliable change in the indirect effect across levels of belonging. These results challenge the assumption that belonging universally amplifies the benefits of teacher support and instead suggest a buffering role of belonging in the Chinese educational context. The findings highlight the importance of providing targeted support to students with low belonging and call for culturally sensitive models of motivation and achievement.

Understanding how groups interact with the task during face-to-face collaborative problem solving in STEM classrooms is central to the task design process of ill-structured tasks. We examine how the presence of hard scaffolds in the design of ill-structured tasks influences groups’ collaborative problem-solving interactions during the engineering task and final task solutions. Our findings indicate that these scaffolds help students structure the task and engage in collaborative problem-solving processes that are usually neglected, which can improve the quality of groups’ final task solutions and give groups a more authentic collaborative problem solving experience in engineering. These findings support the need to include hard scaffolds in the design of ill-structured tasks for collaborative problem solving in higher education STEM classrooms.

This study examines emotion regulation during emotional challenges in mathematics group tasks among teacher education students (N = 62). The data consisted of 87 hours of videotaped group interaction and students’ pre- and post-evaluations of their group experiences. The results show that emotional challenges and regulation varied across groups and tasks. Pre-evaluations were associated with the level of regulation, whereas post-evaluations were associated with the level of challenge but not with regulation. Qualitative case analyses further illustrate group-specific challenges and regulation strategies, including task structuring, awareness raising, encouragement, and social reinforcement. The findings indicate that regulation strategies did not always align with the challenges faced, leading to recurring difficulties within groups. These results have implications for the design and support of collaborative learning: beyond helping students stay on task, groups may benefit from increased awareness of the nature of their challenges and sufficient time to address them through negotiated interaction.

Despite the prevalence of videos and illustrated texts in primary education, there is a limited understanding of how children study and review these materials, and how this impacts their learning performance. This study examined 97 tablet screen recordings, which captured the learning behaviors of fifth graders as they studied illustrated texts and videos in a within-subjects experiment. The screen recordings were coded frame-by-frame to assess the children’s studying rates and learning strategies. These factors, and their connection to prior learning performance data, were analyzed using mixed-effects models and stepwise regression. The results show that videos were studied significantly faster than content-equivalent illustrated texts. Furthermore, this faster studying was associated with better learning outcomes and lower cognitive load. As the children scarcely used the navigation options in the videos, they were mostly studied and reviewed linearly at their designed rate, which was especially beneficial for slower readers. In comparison, both the studying rate and the review strategies were more varied in the illustrated texts. In both conditions, children devoted more time on the slides that had more text content and were positioned early in the presentation. The type of pictures on the slide did not influence time spent. These results indicate that children tend to watch and review videos linearly at their designed rate and sequence, leading to faster studying rates, better learning outcomes, and lower cognitive load compared to illustrated texts. In particular, slow readers benefit from videos.

Dynamic visualization (i.e., animation; video) is often used to improve comprehension of complex materials. This is why it is crucial to facilitate apprehension of dynamic visualization so learners can more easily elaborate a coherent mental representation. In two experiments, we investigated signaling with a pointing task while viewing a video to facilitate apprehension of a dynamic visualization and thus comprehension. Experiment 1 had 78 participants; Experiment 2 had 85 participants. In both experiments, we measured participants’ visuospatial abilities. In both experiments, participants learned about different canine gaits from videos. In Experiment 1, we compared verbal instructions with a pointing task. In Experiment 2, we compared a pointing task with rotation with a pointing task without rotation. We also measured cognitive load. Both experiments showed a positive effect of instructions and pointing tasks (with and without rotation) during video comprehension. Experiment 1 showed a benefit of signaling only for easily apprehensible video and, in contrast, Experiment 2 indicated that pointing tasks were effective regardless of the gait being learned. Our results suggest that a pointing task is not necessarily more useful than instructions alone—learners could misuse the pointing task; also, the pointing task could be effective when the device allows an appropriate use of this task. Another explanation is that learners may have been cognitively fatigued during the first experiment. In this way, signaling could be effective for dynamic visualizations when learners have all their cognitive resources available.

Despite persistent criticism and mixed empirical support, the use of learning styles as a basis for personalising instruction remains prevalent in higher education. This study examined the agreement between two self-report approaches to classifying students’ learning style preferences: a brief self-classification task and the Index of Learning Styles (ILS), a structured self-report questionnaire. A total of 696 undergraduate and master’s students, most of them enrolled in online programmes, completed a brief self-classification task before responding to the ILS. Agreement beyond chance was assessed primarily using Cohen’s Kappa, while McNemar’s test was used as a secondary analysis to examine directional asymmetries in mismatches between the two measures. Across the four dimensions, disagreement rates ranged from 28.30% to 49.60%. Kappa values ranged from − 0.01 to 0.16, indicating little agreement beyond chance, whereas McNemar’s test showed no significant directional asymmetries (p > .05). These findings suggest that a brief self-classification provides only a limited approximation to students’ ILS-based profiles, restricting its usefulness as the sole basis for instructional personalisation. At the same time, the low concordance observed raises questions about the robustness of the ILS as an individual-level diagnostic tool. Importantly, both measures relied on self-report, and the study did not assess learning outcomes or actual study behaviours. The findings should therefore be interpreted as evidence of limited agreement between two self-report approaches to learning style classification, rather than as direct evidence on the effectiveness of matching instruction to learning styles.

Prior meta-analyses and meta-regression studies on peer feedback have consistently observed benefits for providers, but alongside notable and unexplained variability in effect sizes. While some students exhibit minor learning gains from providing feedback, others display significant improvements. We hypothesized that the features of students’ comments may play pivotal roles in the variability of learning by providing. To test these hypotheses, we strategically selected 12 assignments from a comprehensive dataset generated in an online peer feedback platform: six assignments demonstrating weak learning outcomes and six strong ones. The study particularly focused on variation in near-transfer learning rather than improved performance in revised version of the same assignment. Peer comments were strategically selected controlling for comment length and coded on two key dimensions: feedback focus and feedback specificity. Our results revealed that assignments with weak learning outcomes displayed a predominance of lengthy form-focused comments. Conversely, assignments with strong effects were characterized by a prevalence of questions for clarification, suggestions for improvement, and direct revisions concerning content issues. Another notable finding was although the strong cases addressed more content issues than the weak cases did, the difference was not statistically significant, which indicated that the key factor was not simply the greater number of content-focused comments but their specificity. Emphasizing an integrated perspective, we proposed that the ratio of specific content-focused comments to form-focused comments was crucial for identifying situations conducive to student learning through providing feedback. Conceptually, these findings are explained by variation in behavioral engagement and cognitive engagement of the feedback provider.

This quasi-experimental study examines the effects of a literary intervention based on the Quality Talk (QT) approach aimed at promoting dialogic discussions about a moral dilemma. Nine teachers and 104 fourth- and fifth-grade students from Switzerland participated in the study. Over the course of one school year, students in the intervention group read and discussed four literary fiction books in small groups. Teachers received video-based coaching to facilitate the discussions. The subsequent analyses focused on changes in (a) monologic teacher practices, (b) dialogic teacher practices, and (c) the quality of student contributions. Multilevel analyses show that the intervention led to a shift from monologic to dialogic teaching patterns. More specifically, teachers in the intervention group were less likely to ask test questions, control turn-taking, or evaluate students’ responses, but more likely to engage in dialogic practices, as indicated by an increase in the use of talk moves. Due to these shifts in teacher practices, students in the intervention group exhibited higher-quality contributions to literary discussions, as demonstrated by increases in authentic questions, cumulative and exploratory talk, reasoning, and elaborated explanations. Ultimately, these findings provide evidence that QT facilitates a shift toward dialogic patterns, thereby enabling more student-centered and cognitively demanding literary discussions in primary education.

This study investigates recursive cycles of knowledge building-modeling (KBM) supported by knowledge building analytics to foster reflection, model improvement, and systems thinking. A total of 43 grade 5 students in China constructed models of Earth Science systems that served as objects of inquiry in Knowledge Forum. Models were refined as students engaged in knowledge building discourse, using knowledge building analytics to visualize the interconnectedness of ideas and progressively enhance models that link Earth Science phenomena to sustainability. Successive model iterations showed increased modeling sophistication and progression of systems thinking from individual variables to cause–effect relationships, secondary effects, temporal accounts, and predictions of behaviors. Students also demonstrated significant gains in their domain knowledge of Earth Science. Qualitative analyses identified mediating processes and system-level dynamics: (1) idea generation and improvement through collaborative knowledge building discourse, (2) analytics used for meta-reflection and conceptual reorganization, and (3) model construction for system expansion and refinement. These findings illustrate how students appropriated KBM and analytics as epistemic frameworks to engage in recursive design and refinement for systems thinking and knowledge advances. The research demonstrates how computer-supported collaborative learning (CSCL) enriched with model-based inquiry and knowledge building analytics can foster students’ engagement with complex systems.

Teachers’ ability to assess pupils’ knowledge and skills is a crucial component of their daily responsibilities, as accurate diagnoses influence both the support provided to pupils and their learning outcomes. Traditionally, diagnosing pupils’ skills is an individual task, with little consideration given to the potential of collaborative diagnosis. However, collaboration may offer significant benefits, particularly when diagnosing cross-domain skills, such as scientific reasoning, where teachers from different subjects can contribute observations from different classroom situations and integrate their perspectives into a joint diagnosis. Drawing on parallels with the medical field, where collaborative diagnoses have already been shown to improve diagnostic accuracy, potential advantages may also occur in an educational context. Nonetheless, collaboration also comes with difficulties: Distributed information is seldom shared, and collaboration can be ineffective. Thus, using video-based simulations, this study investigates how collaboration, distribution of information, and collaboration support, by means of collaboration scripts, influence diagnostic accuracy – potentially via the application of collaborative diagnostic activities – consequently contributing to the development of diagnostic competences in pre-service teachers. The results indicate that pre-service teachers’ collaborative diagnostic accuracy does not per se exceed individual diagnostic accuracy, but that collaboration support can significantly enhance the added value of collaborative diagnosis. The reason for this seems to be increased information sharing. Thus, this study suggests that successful collaborative diagnoses require collaboration support and that collaboration scripts can help overcome the challenges of collaboration. Future research should explore collaborative diagnosis in different contexts and for other (cross-domain) skills.

In collaborative learning, regulation is crucial for meaningful and effective collective learning. As generative artificial intelligence (AI) becomes readily available to learners as an external, on-demand information resource, regulatory behavior dynamics may shift from information seeking toward evaluating, selecting, integrating, and coordinating externally generated content within group discourse. This study examined the learners’ behavioral dynamics in their cognitive, task, and emotional regulation and investigated the interplay of individual performance and behavior transition patterns against the backdrop of group achievement in an AI-supported collaborative learning environment. A total of 126 undergraduate and graduate students participated in a face-to-face learning task, with access to online AI tools at the learners’ discretion. Audio recordings were transcribed into 16,816 semantic units for content analysis. Using k-means clustering, the research identified four distinct learner types, namely evaluative, curious, expressive, and passive. Lag sequential analysis (LSA) results showed divergent behavioral patterns emerged between high- and low-performance groups and individuals. Group performance level did not warrant similar individual performance levels. The study highlights the complex interplay between individual and group behaviors and their impact on performance, in particular the behaviors of high-performance individuals in low-performing groups and vice versa. Notably, these patterns underscore how group outcomes can hinge on how contributions are regulated and coordinated, rather than on information availability alone. Collectively, the findings suggest that, in AI-supported collaboration, performance is associated with the learner’s capacity to critically appraise and purposely appropriate information among the group members and to align its use with task goals and social emotional coordination. This research offers insights for the facilitators on grouping strategies, differentiated scaffold design for learner types, and intervention considerations for low-performance groups and individuals. Limitations and future research directions are also discussed.

This paper explored the use of a teacher-facing analytics tool designed to support teacher facilitation of collective student inquiry. We studied a Grade 5 science teacher’s implementation of the tool to support students in developing their understanding of electricity. The tool dashboard visualizes the relationship between student discourse and the curriculum by distinguishing shared terms, student-only terms, and conceptual words from the curriculum in the form of visuals including word cloud and word network. Our data included lesson observations (video-recorded), student posts (notes) from an online forum, a science quiz, and interviews. We analyzed video data to understand the teacher’s discourse strategies supported by the tool and assessed students’ online discussions for (i) vocabulary growth in relation to curriculum keywords and (ii) the quality of idea complexity and scientific rigor over time. We also coded interview data to understand participants’ views and perceptions of collaborative learning when supported by the word cloud visualizations. Our findings showed that the teacher used the visualizations to identify emerging ideas aligned with the curriculum and to co-construct promising ideas with students to help advance their inquiry. Our analysis of student data showed that students constructively integrated curriculum keywords from the visualizations to expand idea connections using scientific terms over time. They also reported in interviews that the visualizations promoted their inquiry. The quiz results also showed a significant improvement in their understanding of the topic. We discuss the contributions and implications in relation to teacher-facing learning analytics designed to support collective student inquiry in knowledge building.

Opportunistic collaboration is an emerging and flexible form of collaboration centered on ideas and open interaction structures, increasingly adopted in collaborative learning contexts, especially in knowledge-building environments. Although prior studies have discussed its potential benefits and design strategies, little is known about how students actually engage in opportunistic collaboration in an authentic knowledge-building context. Adopting a process-oriented perspective, this study investigated how different students participate in opportunistic collaboration and how they experience this collaboration. The study was conducted in a graduate-level learning sciences course with 24 master’s students involved in a semester-long knowledge-building inquiry. Guided by distance-shortening strategies that reduced both physical and idea distance, the learning environment supported students’ movement and adaptive collaboration. Using an explanatory mixed-methods design, quantitative data included classroom movement frequency, interaction behaviors, and 314 online notes analyzed through two-step cluster analysis, correlation analysis, and content analysis. Qualitative data were collected through semistructured interviews and analyzed using thematic analysis. The results identified two distinct participation patterns. One group of students demonstrated higher engagement with frequent movement and deeper idea development, while the other showed lower engagement with limited movement and more surface-level idea contributions. These patterns were further interpreted as two qualitatively different modes of participation: a cognitively driven mode and a contextually driven mode. The findings reveal that opportunistic collaboration unfolds as a differentiated process shaped by students’ epistemic orientations, motivations, and prior learning habits. This study provides a process-oriented insight into the nature of opportunistic collaboration and offers implications for designing pedagogical strategies in knowledge-building environments.

Recent disruptions have prompted scholarly inquiry into whether research on knowledge building (KB) represents transient disturbances or durable transformation at the disciplinary level. This study offers a phase-specific synthesis of KB research from 2014 to 2025, divided into three periods: a pre-pandemic baseline (2014–2019), a pandemic-induced adaptation phase (2020–2022), and a post-2023 era marked by integration of generative artificial intelligence (GenAI) and learning analytics. Employing Preferred Reporting Items for Systematic Reviews and Meta-analyses (PRISMA)-guided retrieval and explicit KB inclusion criteria, 306 KB-core journal articles indexed in the Social Sciences Citation Index (SSCI) were analyzed through science-mapping and systematic coding of research designs, contexts, and platforms. Findings reveal structural shifts in publication volume around 2020 and 2023, followed by sustained post-2023 acceleration. Considering publication lag, this period is interpreted as temporally buffered, with trends treated as early signals rather than definitive outcomes. Core KB constructs remain stable across phases, indicating continuity at the level of epistemic mechanisms. Post-2023 developments show selective integration of learning analytics and GenAI into the KB framework, rather than replacement of established constructs, with explicit AI integration showing a regionally concentrated early adoption pattern in East Asia. Collaboration networks expand with increased connectivity and decreased assortativity, suggesting growth with coherence. Collectively, these findings indicate that recent changes do not constitute a paradigmatic shift but reflect constraint-driven reorganization under enduring epistemic commitments. This reorganization entails selective integration of emerging technologies with knowledge-building principles while raising tensions related to transparency, interpretability, and epistemic agency. This study contributes a field-level framework for distinguishing technological perturbations from enduring epistemic structures, highlighting the need for transparency and governance to support cumulative inquiry in AI-mediated computer-supported collaborative learning (CSCL).