Bora Başaran is an Associate Professor of German Language Education at Anadolu University, Türkiye. His research focuses on artificial intelligence in education, technology-enhanced language learning, German as a foreign language, teacher education, and AI-supported assessment and feedback. His recent work examines human–AI collaboration in language education, AI attitudes and higher-order thinking, automated language assessment, and the pedagogical integration of generative AI. He also develops and evaluates AI-supported learning environments for speaking, writing, and research mentoring, with particular attention to teacher agency, learner autonomy, privacy, and responsible AI use. Alongside his research, he teaches undergraduate and graduate courses on German language teaching, digital pedagogy, and artificial intelligence in language education.
Critical cultural awareness is a central goal of language education, yet the authentic socio-political discourse through which target-language societies debate their own social categories is still rarely used as classroom material. One reason is that the cultural framing embedded in such discourse is difficult for learners to perceive without analytic support. This paper reports the design and pre-classroom evaluation of a browser-based task environment in which learners of German annotate excerpts from family-policy documents from Turkey, Germany, Switzerland and Austria through three learner-facing lenses. They justify each interpretation with reference to the wording of the text and receive question-based rather than answer-based feedback. The feedback layer is implemented deterministically as a rule-based Wizard-of-f-Oz version of a prompt contract, allowing the form of feedback to be evaluated independently of any particular language model. The evaluation combines an 84-item simulated-learner stress test with sessions involving five practising teachers of German in two arrangements, co-located and independent, as well as twenty held-out learner-voice justifications that the rules had not previously encountered. Six failure modes are documented. Two are especially consequential, showing that lexical overlap between a justification and a selected span fails in both directions as a proxy for interpretive anchoring. We also describe a verification procedure for unsupervised browser-based evaluation in which each submitted session is replayed against the frozen artefact itself.
Dr. Chuang Wang is Chair Professor of Educational Research and Evaluation at the Faculty of Humanities of The Education University of Hong Kong (EdUHK), Director of the Center for Higher Education Leadership and Policy Studies at EdUHK, former Chair of the Department of Educational Leadership at the University of North Carolina at Charlotte, and former Dean of the Faculty of Education at the University of Macau. His research interests include self-efficacy beliefs, self-regulation, and use of AI in large data analysis. He has published 8 books, 22 book chapters, and over 200 peer-reviewed journal articles. He was listed as one of the world masters in English language education by China Daily and top 2% scientists worldwide by Stanford University. Professor Wang has also received awards in research, teaching, and service at the university level and academic associations such as American Educational Research Association, the U.S. Academy of Educational Leadership, the Association for Educational Communications and Technology, and the Chinese American Educational Research and Development Association. He served as the Editor-in-Chief of the Educational Research & Development Journal and the Journal of Applied Educational and Policy Research. He is Associate Editor of Psychology in the Schools and Sage Open.
AI literacy is increasingly recognized as an essential competency in education. Yet current AI literacy initiatives overwhelminglytarget end-learners or in-service teachers, neglecting the foundationallayer of the education pipeline that starts from higher education facultywho design teacher preparation curricula, to the teachers they train,and ultimately to the K–12 students those teachers instruct. If the population at the source of the pipeline remains underserved, the entirepipeline stalls. Understanding what these current and future faculty already know about AI and computational tools, and identifying the barriers they face, is therefore a critical leverage point for systemic change.This work explores what education faculty and graduate students currently know about AI and related computational technologies throughliterature review and user study.
Jowati Juhary is an academic at the Language Centre of the National Defence University of Malaysia (UPNM) in Kuala Lumpur. She is currently the Dean of the Language Centre. She has a Bachelor of Education in Teaching English as a Second Language (B.Ed (TESL)), an MA in English Language Studies, and a PhD in Governance and Development. Her research centres on four core areas: first, educational technology; second, general teaching methods, namely pedagogy and military pedagogy; third, higher education; and fourth, language education related to English language teaching. As early as 2005, she published her first relevant study, which argued for the implementation of e-learning in Malaysia’s military academies. Over the nearly two decades since then, she has continuously published relevant articles, consistently exploring the impacts that technology has brought to teaching in military learning environments. Her specific research covers a wide range of topics, including the blended learning model; the development of e-learning content suitable for online teaching; and the learning management systems that universities use to uniformly manage courses, assignments, and grades, from the small-scale use of social media as a review tool for students to the application of blended teaching and the HyFlex model in the context of Industry 4.0 and the post-pandemic era, all of which fall within her research scope.
Artificial Intelligence (AI) is repeatedly discussed in the global higher education sector, and it is widely recognized that it will drive a thorough transformation of higher education. Local universities in Malaysia have only recently begun to catch up with this wave of transformation, attempting to access and deploy AI-related applications. This paper collates all relevant international literature published over the seven-year period from 2017 to 2023, specifically extracts the most frequently discussed and recurring common content, and organizes it around three core directions: first, what specific scenarios AI can be applied to in universities; second, what implementation strategies universities need to successfully put AI into practical use; and third, what deeper impacts the introduction of AI will bring to the entire higher education system. After sorting through all the literature, several clear conclusions can be drawn. First, the three types of AI applications most discussed in all studies are adaptive learning platforms, automatic assessment tools, and intelligent tutoring systems that can answer students’ questions and provide guidance at any time. Second, to successfully deploy AI on a large scale in universities, almost all studies repeatedly mention three unavoidable necessary conditions, including special investment must be made to build infrastructure; universities must not work behind closed doors; and digital literacy training must be provided for educators, administrative staff and all students in the university, so that everyone is able to use these new tools. Finally, the three types of risks most mentioned in all studies are the staffing structure of original teaching and administrative positions in universities will be impacted; various AI tools for students may have unfair algorithmic biases; and the vast majority of universities lack sound management frameworks, making it impossible to supervise the use of AI in compliance with ethical norms. At the end of the paper, specific and directly implementable suggestions are put forward for top management of Malaysian higher education, and a preliminary local research plan is also provided.
Dr. Fernando Martin was a Fulbright Fellow in Music in the USA and studied at the University of Washington, University of California, and Tanglewood Music Center in Massachusetts attending lectures with Yo-Yo Ma and John Williams. He later received the AHRB fellowship from the UK government and studied at King's College London (UK), Conservatorio Superior in Madrid (Spain), and IRCAM Centre Pompidou in Paris (France). He has published articles on Igor Stravinsky and has given talks at McGill University (Canada), University of Birmingham (UK), and the University of Cambridge.
While research has established the importance of a supportive team environment, motivation, and collaborative dynamics in learning, a fundamental question remains: once a safe team environment is established, what procedural mechanisms translate team potential into innovation, effective problem-solving, and meaningful learning?
This paper proposes a unified framework that synthesizes two approaches—Google's collaborative learning practices and the Kaizen Event method—and introduces an audit to determine whether educators' assessments align.
The first approach, Google's Project Aristotle, identified psychological safety and meaning as central to team effectiveness. Yet this finding does not tell educators how to design assessments that support such conditions. The Kaizen Event offers a six-phase process—Problem Definition, Data Collection, Root Cause Analysis, Solution Generation, Implementation Planning, and Verification—with agreement between phases. It is a method used in industry to promote continuous learning through structured group problem-solving. Together, these approaches offer insights into group dynamics and procedural mechanisms that can be adapted for assessment design.
The framework proposed converts these insights into 36 conditions, each a requirement an assessment must meet to enable high-quality teamwork. The conditions fall into four categories: the procedures the team follows (mechanisms), the records and products it generates (documents and artefacts), the divergence–convergence cycle that promotes high performance (motor), and how it handles its own dynamics (rules). Safety is not a fifth category: as Google suggests, it is the condition that enables the other four.
The proposed audit shows alignment against this framework, and can be supported by Epistemic Network Analysis (ENA) and a large language model (LLM). Together they analyze the team's discourse and detect emergent behaviors and constructive conflict.
We illustrate the framework with examples from music, computer engineering and business, and suggest future research. Ultimately, this paper offers tools for designing assessments that build emergent team learning.
Dr Stella Stefany is a Lecturer in the School of Management at RMIT University, Melbourne, where her research and teaching centre on serious games, game-based learning and playful pedagogy in business education. Her doctoral research examined the design and curricular embedding of Blackout, a design-thinking serious game in a first-year business course. The game won second place in the International Serious Games Competition at the 2024 European Conference on Games Based Learning. She is Principal Investigator of Playful Pedagogy for Transition Support, a longitudinal design-based research programme on first-year student transition.
Her work addresses the gap between how educators and students engage with learning technologies, positioning games as tools that extend, rather than replace, educators' pedagogical judgement. Previously, she was Head of Department and Senior Lecturer in Online Learning Communication Science at Universitas Pelita Harapan, Indonesia, and she received Turnitin's Global Innovator Award in 2019. Her teaching has earned several Outstanding Good Teaching Scale awards at RMIT, and her work has featured in The Conversation, and SBS. She values interdisciplinary and international collaboration.
Debates about educational technology tend to ask whether a tool works. This talk argues that we have been asking the wrong question. A better one is what conditions must be in place for any learning technology to produce learning. It draws on a research programme on playful pedagogy and first-year transition at RMIT University, and in particular on forthcoming work on Blackout, a crisis-management serious game played by more than 1,200 business students. From that work it proposes three enablers of experiential learning. The first is safety-driven motivation: consequences that feel real inside the game but bounded outside it, so students can risk being wrong. The second is immersive engagement: absorption that spills into conversation with peers, even in a single-player game. The third is educator enthusiasm: visible, energetic facilitation that turns play into reflection. Together, these explain how a intentionally designed serious game moves students through Kolb's experiential learning cycle. In each case the enabler is human and relational, not technical. As generative AI makes polished outputs easy to produce, these processes of judgement become more valuable, not less. The talk closes with open questions for the field: designing for online and large cohorts, developing sessional educators, and moving GenAI policy from detection towards learning that cannot be delegated.