InsPIRE NTU

We are pleased to announce NTU’s Annual Learning and Teaching Conference: From Good to Great 2026 (13–14 May 2026) at NTU, Singapore. This year’s theme is “What’s higher about higher education? Thriving together in an age of artificial intelligence.” We invite educators, researchers, and practitioners to submit proposals for papers or roundtable discussions, or to register to attend (no registration fee). Do share this with colleagues who may be interested. More details on keynote speakers, workshops, and the full programme will be shared soon. https://lnkd.in/gejiqttz

𝗔𝗿𝗲 𝘆𝗼𝘂𝗿 𝗮𝘀𝘀𝗲𝘀𝘀𝗺𝗲𝗻𝘁𝘀 𝗿𝗲𝗮𝗹𝗹𝘆 𝗺𝗲𝗮𝘀𝘂𝗿𝗶𝗻𝗴 𝘄𝗵𝗮𝘁 𝘆𝗼𝘂 𝘁𝗵𝗶𝗻𝗸 𝘁𝗵𝗲𝘆 𝗮𝗿𝗲? 🤔 A student submits excellent code but conceals terrible coding practices. A group project earns top marks, but one student did all the work whilst others contributed nothing. Sound familiar? This new open-access article by Tim Fawns, Phillip Dawson, and David Boud addresses a critical challenge in higher education: the gap between what we 𝘪𝘯𝘵𝘦𝘯𝘥 to assess and what we 𝘢𝘤𝘵𝘶𝘢𝘭𝘭𝘺 measure. 𝗧𝗵𝗲 𝟰𝗣𝘀 𝗙𝗿𝗮𝗺𝗲𝘄𝗼𝗿𝗸 introduces four types of evidence in assessment: • 𝗣𝗿𝗼𝗱𝘂𝗰𝘁 – tangible outputs like essays or projects • 𝗣𝗿𝗼𝗰𝗲𝘀𝘀 – the steps and strategies used • 𝗣𝗲𝗿𝗳𝗼𝗿𝗺𝗮𝗻𝗰𝗲 – real-time demonstration of skills • 𝗣𝗿𝗮𝗰𝘁𝗶𝗰𝗲 – situated application in authentic contexts 𝗞𝗲𝘆 𝗶𝗻𝘀𝗶𝗴𝗵𝘁𝘀: • Simply mapping assessment tasks to learning outcomes isn’t enough – we need to scrutinise the 𝘵𝘺𝘱𝘦 of evidence each assessment generates • Over-reliance on products (like essays and exams) can mask whether students have truly achieved the intended capabilities • Different proxies reveal different aspects of learning – combining them strategically strengthens assessment validity • Common pitfalls include 𝘮𝘪𝘴𝘮𝘢𝘵𝘤𝘩𝘪𝘯𝘨 (wrong proxy for the outcome), 𝘴𝘭𝘪𝘱𝘱𝘢𝘨𝘦 (assessing one proxy whilst intending another), and 𝘴𝘱𝘪𝘭𝘭𝘢𝘨𝘦 (unintentionally assessing additional proxies) The framework offers a practical tool for educators to interrogate their assessment practices and design more credible, balanced assessment systems – without requiring specialist expertise in psychometrics. Read the full open-access article: https://lnkd.in/g_Sn_Pdz #Assessmentntu #SoLntu #TandLntu #AcademicIntegrity #AuthenticLearning #Assessment #ScienceOfLearning #InsPIREntu

🚨 𝗪𝗵𝘆 𝗔𝗜 𝗗𝗲𝘁𝗲𝗰𝘁𝗼𝗿𝘀 𝗠𝗮𝘆 𝗗𝗼 𝗠𝗼𝗿𝗲 𝗛𝗮𝗿𝗺 𝗧𝗵𝗮𝗻 𝗚𝗼𝗼𝗱 Are you relying on AI detection tools to catch unauthorised AI use in student assessments? A compelling new open-access article in the 𝘑𝘰𝘶𝘳𝘯𝘢𝘭 𝘰𝘧 𝘏𝘪𝘨𝘩𝘦𝘳 𝘌𝘥𝘶𝘤𝘢𝘵𝘪𝘰𝘯 𝘗𝘰𝘭𝘪𝘤𝘺 𝘢𝘯𝘥 𝘔𝘢𝘯𝘢𝘨𝘦𝘮𝘦𝘯𝘵 argues that AI detectors are fundamentally flawed and should not be used in educational settings. 📌 𝗞𝗲𝘆 𝘁𝗮𝗸𝗲𝗮𝘄𝗮𝘆𝘀: • 𝗨𝗻𝘃𝗲𝗿𝗶𝗳𝗶𝗮𝗯𝗹𝗲 𝗿𝗲𝘀𝘂𝗹𝘁𝘀: Unlike plagiarism detection, AI detectors rely on probabilistic estimates that cannot be verified in real-world conditions where we don’t know the true origin of text • 𝗙𝗮𝗹𝘀𝗲 𝗱𝗶𝗰𝗵𝗼𝘁𝗼𝗺𝘆: The assumption that text is either entirely human-written or AI-generated ignores how students actually work 𝘸𝘪𝘵𝘩 AI, not just 𝘣𝘺 AI • 𝗣𝗿𝗼𝗰𝗲𝗱𝘂𝗿𝗮𝗹 𝗳𝗮𝗶𝗿𝗻𝗲𝘀𝘀 𝗰𝗼𝗻𝗰𝗲𝗿𝗻𝘀: Detection scores don’t meet the balance-of-probabilities standard required for academic integrity investigations • 𝗖𝗼𝗻𝗳𝗶𝗿𝗺𝗮𝘁𝗶𝗼𝗻 𝗯𝗶𝗮𝘀: Common validation methods — checking multiple detectors, hunting for ‘AI hallmarks’, comparing to past writing — reinforce assumptions rather than provide independent verification • 𝗖𝗹𝗶𝗺𝗮𝘁𝗲 𝗼𝗳 𝘀𝘂𝘀𝗽𝗶𝗰𝗶𝗼𝗻: Surveillance approaches undermine trust and may lead innocent students to be wrongly accused The authors make a strong case that institutions need to shift from detection and enforcement to 𝘳𝘦𝘵𝘩𝘪𝘯𝘬𝘪𝘯𝘨 𝘢𝘴𝘴𝘦𝘴𝘴𝘮𝘦𝘯𝘵 𝘥𝘦𝘴𝘪𝘨𝘯 that acknowledges AI’s role in contemporary learning. Worth a read if you’re navigating the complexities of GenAI in your teaching practice. The article is open access and available at the link below. https://lnkd.in/gYuGbBr3 💭 What are your experiences with AI detection tools? Have you moved towards alternative assessment strategies? #GenAIntu #TandLntu #Assessmentntu #academicintegrity #resourcentu #GenAI

📚 𝗡𝗲𝘄 𝗥𝗲𝗽𝗼𝗿𝘁: 𝗢𝗘𝗖𝗗 𝗗𝗶𝗴𝗶𝘁𝗮𝗹 𝗘𝗱𝘂𝗰𝗮𝘁𝗶𝗼𝗻 𝗢𝘂𝘁𝗹𝗼𝗼𝗸 𝟮𝟬𝟮𝟲 🤖 The latest OECD flagship report explores how generative AI can transform teaching and learning, but only when used with clear pedagogical intent. 𝗞𝗲𝘆 𝗶𝗻𝘀𝗶𝗴𝗵𝘁𝘀 𝗳𝗼𝗿 𝘂𝗻𝗶𝘃𝗲𝗿𝘀𝗶𝘁𝘆 𝗲𝗱𝘂𝗰𝗮𝘁𝗼𝗿𝘀: • 𝘗𝘦𝘳𝘧𝘰𝘳𝘮𝘢𝘯𝘤𝘦 ≠ 𝘓𝘦𝘢𝘳𝘯𝘪𝘯𝘨: Students using generic GenAI tools improved task performance by up to 48%, but performed 17% worse once AI access was removed—highlighting the risk of creating a “mirage of false mastery” rather than genuine learning • 𝘗𝘶𝘳𝘱𝘰𝘴𝘦-𝘣𝘶𝘪𝘭𝘵 𝘵𝘰𝘰𝘭𝘴 𝘮𝘢𝘵𝘵𝘦𝘳: Educational AI designed as learning partners—using Socratic questioning and requiring students to explain, critique, or build upon AI outputs—can strengthen conceptual understanding and critical thinking • 𝘛𝘦𝘢𝘤𝘩𝘦𝘳 𝘢𝘶𝘵𝘰𝘯𝘰𝘮𝘺 𝘮𝘶𝘴𝘵 𝘳𝘦𝘮𝘢𝘪𝘯 𝘤𝘦𝘯𝘵𝘳𝘢𝘭: Whilst GenAI can reduce lesson planning time by 31%, uncritical automation of teaching tasks risks undermining professional judgement and pedagogical expertise • 𝘚𝘺𝘴𝘵𝘦𝘮𝘪𝘤 𝘢𝘱𝘱𝘳𝘰𝘢𝘤𝘩 𝘳𝘦𝘲𝘶𝘪𝘳𝘦𝘥: Digital transformation succeeds only when technology is embedded within coherent, system-wide reform, and not treated as isolated tools or pilots The report emphasises that human judgement, feedback, and oversight must remain at the centre of AI use in education. 𝘙𝘦𝘢𝘥 𝘵𝘩𝘦 𝘧𝘶𝘭𝘭 𝘳𝘦𝘱𝘰𝘳𝘵: https://lnkd.in/ekQjriRM 💭 How are you approaching GenAI integration in your courses? Are you designing AI interactions that promote deep learning rather than task completion? #GenAIntu #TELntu #TandLntu #resourcentu #GenAI #edtech

𝗛𝗮𝗽𝗽𝘆 𝗖𝗵𝗶𝗻𝗲𝘀𝗲 𝗡𝗲𝘄 𝗬𝗲𝗮𝗿 𝗳𝗿𝗼𝗺 𝗜𝗻𝘀𝗣𝗜𝗥𝗘! 🍊🧧 As we prepare to welcome the 𝗬𝗲𝗮𝗿 𝗼𝗳 𝘁𝗵𝗲 𝗛𝗼𝗿𝘀𝗲, we wish all our academic colleagues a season filled with prosperity, good health, and boundless vigour. The horse is a symbol of energy, perseverance, and swift progress - qualities that resonate deeply with the academic journey. May this year bring you the stamina to pursue new breakthroughs in your research and the creative spirit to inspire your students in the classroom. We hope you take this opportunity to rest, recharge with loved ones, and return refreshed for the semester ahead. #InsPIREsharentu #wellbeingntu

𝗛𝗮𝗽𝗽𝘆 𝗩𝗮𝗹𝗲𝗻𝘁𝗶𝗻𝗲’𝘀 𝗗𝗮𝘆! 💖 They say that teaching is a work of heart, and we certainly agree. Today, we simply want to celebrate the passion, patience, and dedication you bring to the university every day. Whether it is staying back to answer one last question, redesigning an assessment to be more authentic, or finding new ways to make complex topics click, your commitment to student success does not go unnoticed. Thank you for putting the “love” in the love of learning! 🍎 #wellbeingntu #InsPIREsharentu #academicintegrity #teachinglife #InsPIREntu

We often treat student writing like private property: fenced off, owned by a single "solitary genius," and strictly policed. 🚧 But does that fence still make sense in a networked world? In a thought-provoking new piece, Eric Martinsen argues that our current anxiety about AI stems from a clash between this old "enclosure" model and the reality of how thinking actually happens today. When we view writing solely as a product to be owned, AI feels like a threat to integrity. The institutional reflex is often to build taller fences—more surveillance, more detection tools, and an "algorithmic panopticon" that replaces curiosity with compliance. Martinsen suggests a powerful shift: moving from 𝗜𝗻𝘁𝗲𝗹𝗹𝗲𝗰𝘁𝘂𝗮𝗹 𝗘𝗻𝗰𝗹𝗼𝘀𝘂𝗿𝗲 to 𝗜𝗻𝘁𝗲𝗹𝗹𝗲𝗰𝘁𝘂𝗮𝗹 𝗦𝘁𝗲𝘄𝗮𝗿𝗱𝘀𝗵𝗶𝗽. Instead of asking "Is this yours?", we should be asking: • 𝗛𝗼𝘄 𝘄𝗮𝘀 𝘁𝗵𝗶𝘀 𝘀𝗵𝗮𝗽𝗲𝗱? • 𝗪𝗵𝗮𝘁 𝗶𝗻𝗳𝗹𝘂𝗲𝗻𝗰𝗲𝘀 (𝗶𝗻𝗰𝗹𝘂𝗱𝗶𝗻𝗴 𝗔𝗜) 𝗱𝗶𝗱 𝘆𝗼𝘂 𝗱𝗿𝗮𝘄 𝗼𝗻? • 𝗪𝗵𝘆 𝗱𝗼 𝘆𝗼𝘂 𝘀𝘁𝗮𝗻𝗱 𝗯𝗲𝗵𝗶𝗻𝗱 𝘁𝗵𝗲𝘀𝗲 𝗰𝗹𝗮𝗶𝗺𝘀? This approach doesn't lower standards; it relocates them. It values judgment over mere output and makes the "messy" process of thinking visible again. If you are rethinking how to assess writing without turning your classroom into a police state, this is a must-read. Read the full article here: https://lnkd.in/gEsU7Mgw #GenAIntu #Assessmentntu #academicintegrity #SoLntu #TandLntu #InsPIREsharentu #GenAI

We invest heavily in improving feedback, focusing on skills, rubrics, and models. But what if the single biggest factor is one we're not even discussing: the specific, unwritten 𝗰𝘂𝗹𝘁𝘂𝗿𝗲 of our discipline? This fascinating seminar from Deakin University's CRADLE, "Transforming feedback cultures", explores this very idea. Led by Prof. Margaret Bearman, it draws on a 5-year study in medical training to make a compelling case that feedback isn't a generic skill. Instead, it's powerfully shaped by the specific practices, values, and (most importantly) the 𝘂𝗻𝗰𝗲𝗿𝘁𝗮𝗶𝗻𝘁𝗶𝗲𝘀 of a field. A few key takeaways for all educators: 💡 𝗧𝗵𝗲 '𝗣𝗿𝗼𝗯𝗹𝗲𝗺 𝗼𝗳 𝗖𝗲𝗿𝘁𝗮𝗶𝗻𝘁𝘆': Feedback often gravitates only to what is certain (e.g., technical skills) while avoiding complex, uncertain areas (like communication or decision-making under pressure) [37:56]. 💡 '𝗣𝗮𝘁𝗰𝗵𝗶𝗻𝗴 𝗧𝗼𝗴𝗲𝘁𝗵𝗲𝗿' 𝗣𝗿𝗼𝗴𝗿𝗲𝘀𝘀: When formal feedback is misaligned with the culture, learners are left to 'patch together' an incomplete picture of their performance from scraps of information [22:56]. 💡 𝗖𝘂𝗹𝘁𝘂𝗿𝗲 𝗼𝘃𝗲𝗿 𝗦𝗸𝗶𝗹𝗹𝘀: The seminar contrasts surgery (collocated, frequent task feedback) with intensive care (uncertain outcomes, tacit validation) to show how culture dictates feedback [16:15]. It's a long-form, rigorous discussion that challenges us to look beyond individual 'feedback literacy' and start analysing the 𝘴𝘺𝘴𝘵𝘦𝘮𝘴 we work within. What are the unwritten rules of feedback in your discipline? 🤔 Explore the full seminar here: https://lnkd.in/gTz9tjmK #Assessmentntu #TandLntu #ScienceofLearning #Assessment #InsPIREsharentu #resourcentu

The rise of Generative AI brings immense potential, but recent developments highlight the critical importance of safety guardrails and ethical considerations in these technologies. Singapore authorities are currently engaging with X (formerly Twitter) following reports that its AI chatbot, Grok, has been generating non-consensual sexually explicit deepfakes. This case serves as a vital discussion point for us as educators, particularly regarding digital literacy, AI ethics, and the wellbeing of our university community. 𝗛𝗲𝗿𝗲 𝗮𝗿𝗲 𝘀𝗼𝗺𝗲 𝗸𝗲𝘆 𝘁𝗮𝗸𝗲𝗮𝘄𝗮𝘆𝘀 𝗳𝗿𝗼𝗺 𝘁𝗵𝗲 𝗿𝗲𝗰𝗲𝗻𝘁 𝗿𝗲𝗽𝗼𝗿𝘁𝘀: ⚠️ 𝗥𝗲𝗴𝘂𝗹𝗮𝘁𝗼𝗿𝘆 𝗔𝗰𝘁𝗶𝗼𝗻: The Infocomm Media Development Authority (IMDA) is in talks with the platform to curb the spread of harmful content, while neighbouring countries like Indonesia and Malaysia have already moved to block access to the tool. ⚠️ 𝗦𝗮𝗳𝗲𝘁𝘆 𝗚𝗮𝗽𝘀: Despite updates, concerns persist about the ease with which users can bypass safety filters to create non-consensual intimate images, raising questions about current industry standards for AI safety. ⚠️ 𝗙𝘂𝘁𝘂𝗿𝗲 𝗦𝘂𝗽𝗽𝗼𝗿𝘁: The upcoming Online Safety Commission (OSC), set to be established by June 2026, will provide a clearer avenue for victims of online harms to seek redress. As we integrate GenAI into our teaching and learning, cases like this underscore the need to equip our students not just with technical skills, but with a strong ethical framework for responsible AI use. ❓ 𝗛𝗼𝘄 𝗮𝗿𝗲 𝘆𝗼𝘂 𝗮𝗱𝗱𝗿𝗲𝘀𝘀𝗶𝗻𝗴 𝘁𝗵𝗲 𝗲𝘁𝗵𝗶𝗰𝗮𝗹 𝗿𝗶𝘀𝗸𝘀 𝗼𝗳 𝗚𝗲𝗻𝗔𝗜 𝗶𝗻 𝘆𝗼𝘂𝗿 𝗰𝘂𝗿𝗿𝗶𝗰𝘂𝗹𝘂𝗺? Read the articles here: 🔗 https://lnkd.in/gWAyE-m4 🔗 https://lnkd.in/g7sMCrs6 #GenAIntu #wellbeingntu #academicintegrity #GenAI #edtech #InsPIREsharentu

Are you a 𝗖𝘆𝗯𝗼𝗿𝗴, a 𝗖𝗲𝗻𝘁𝗮𝘂𝗿, or a 𝗦𝗲𝗹𝗳-𝗔𝘂𝘁𝗼𝗺𝗮𝘁𝗼𝗿? 🤖🧠 As we integrate Generative AI into our teaching and research workflows, the debate often centres on “using it” versus “not using it”. However, a fascinating working paper from Harvard Business School suggests the reality is far more nuanced—and the mode of collaboration we choose has profound implications for the future of expertise. In a study of 244 management consultants, researchers identified three ways professionals actually work with AI. They identified three distinct “modes” of work, each leading to very different learning outcomes: • 𝗖𝘆𝗯𝗼𝗿𝗴𝘀 (𝗙𝘂𝘀𝗲𝗱 𝗖𝗼-𝗖𝗿𝗲𝗮𝘁𝗶𝗼𝗻): These professionals integrated AI continuously throughout their workflow. They engaged in a “jagged frontier” of constant back-and-forth conversation. The result? They developed entirely new AI-related capabilities (“Newskilling”). • 𝗖𝗲𝗻𝘁𝗮𝘂𝗿𝘀 (𝗗𝗶𝗿𝗲𝗰𝘁𝗲𝗱 𝗖𝗼-𝗖𝗿𝗲𝗮𝘁𝗶𝗼𝗻): These users maintained strict control, delegating specific sub-tasks to AI while retaining the cognitive “steering wheel”. By doing so, they actually deepened their existing domain expertise (“Upskilling”). • 𝗦𝗲𝗹𝗳-𝗔𝘂𝘁𝗼𝗺𝗮𝘁𝗼𝗿𝘀 (𝗔𝗯𝗱𝗶𝗰𝗮𝘁𝗲𝗱 𝗖𝗼-𝗖𝗿𝗲𝗮𝘁𝗶𝗼𝗻): This group handed off the entire problem-solving process to the AI. While efficient, the study suggests they missed out on both domain learning and AI skill acquisition (“No skilling”). 𝗪𝗵𝘆 𝘁𝗵𝗶𝘀 𝗺𝗮𝘁𝘁𝗲𝗿𝘀 𝗳𝗼𝗿 𝘂𝘀 𝗮𝘁 𝗡𝗧𝗨: This framework forces us to ask critical questions about our pedagogy. If our assessments encourage “Self-Automation”, are we robbing students of the chance to build expertise? Conversely, how can we design learning activities that mould students into capable Centaurs and Cyborgs? Read the full working paper here: https://lnkd.in/gY9n6Cyq #GenAIntu #SoLntu #TandLntu #futureofwork #GenAI #InsPIREsharentu #ScienceofLearning