🚀 Thrilled to Share Big News! 🚀At SEFI24, @ScottDanie77858 and @elementalish had the honour of accepting the EJEE Best Paper Award for the 2023 volume on behalf of our dedicated team! 🏆 The European Journal of Engineering Education, ranked 115th among 1,543 education journals, recognised our work, which currently stands as the 4th most-read paper in EJEE history with over 30,461 downloads. 📊 EJEE has an amazing community, passionate and supportive! Our study https://t.co/IN9hInAQki was one of the pioneering and most comprehensive analyses of the risks and benefits of ChatGPT in higher education assessments, setting the stage for much of today’s thinking on GenAI assessment security.
But we didn’t stop there—our latest research https://t.co/wRGWuMjCbe goes beyond ChatGPT, exploring other AI models and delivering even deeper insights.✨
🔍 Key Highlights:
Table 10: Your ultimate guide to short- and long-term assessment security and integration strategies.
Supplementary Material: Practical tools to assess the risks in your own practices and help you redesign assessments for the future.
Scott and Peter also led a workshop at SEFI24 on this risk assessment and assessment redesign process, and the feedback was incredible—with many attendees calling it one of the best experiences of the conference! 🙌
A huge thank you to the @SEFI_HQ - European Society for Engineering Education community for your support and recognition of our work. We’re excited to continue pushing the boundaries of educational research together! 🚀🎉
With great appreciation from Sasha Nikolic, @ScottDanie77858 , @RezwanulHaqueAU , Marina Belkina, @gmjally , Sarah Grundy (PhD Chemical Engineering), Sarah Lyden, @elementalish and Caz Sandison @uoweis
If you are interested in our further research within the GenAI space, please visit: https://t.co/RCKNOcOKWq
@scientistswrite ChatGPT, Copilot, Gemini, SciSpace and Wolfram versus higher education assessments: an updated multi-institutional study of the academic integrity impacts of Generative Artificial Intelligence (GenAI) on assessment, teaching and learning in engineering
https://t.co/U25Tyu5mp8
Clear evidence of how major models have improved responses to a wide range of assessments across multiple institutions, your security options, and how to integrate #GenAI into future tasks. Based in #STEM but applicable to all disciplines and levels.
Key #ChatGPT, SciSpace and Wolfram GPT, #Gemini and #Copilot vs Higher Education #Assessment insights (Open Access) - https://t.co/wRGWuMjCbe
- Table 10 offers a practical guide to integration opportunities and outlines both short-term and long-term assessment security options for each type of assessment.
- There are short and long-term assessment security options available to support assessment integrity, but long-term, integration is the only option for non-supervised assessment
- Each LLM excels in different areas, enhancing student support across various assessment types.
- Assessment security depends on the assessment type rather than the subject matter.
- Image uploads in ChatGPT-4o are improving significantly, though still a work in progress.
- Wolfram GPT generally provides accurate answers to calculation questions.
- Combining SciSpace GPT with ChatGPT-4 effectively mitigates hallucination issues, engages with recent literature, and enhances productivity in research writing.
- Project work continues to be an excellent opportunity for GenAI integration.
- Now, more than ever, it is essential to focus on learning objectives that target skills beyond #GenAI capabilities.
- If you tried ChatGPT-3.5 and found it lacking, it’s time to retest with GPT-4o
Super excited to share an update and expansion of the work we published in 2023. Check out our schema for assessing and controlling risks, and identifying opportunities for #GenAI integration. https://t.co/U25Tyu5mp8
Excited that our paper "#ChatGPT vs Engineering #Education assessment" in @EJEE_Editor is now available (free)! We evaluate #AI capacity to complete assessments from 10 subjects x 7 institutions to test and highlight strengths, weaknesses and opportunities.
But even more, we introduce a schema that provides practical guidance on managing #AssessmentIntegrity risks and opportunities to integrate #GenAI to enhance learning.
Very happy to share our new paper providing clear, detailed evidence of the dramatic need to redesign and reconceptualise #assessment for the #GenAI age. https://t.co/Tj94xxGsgH
In this video we summarise the threats & opportunities of #AI in #assessment (https://t.co/AYxnqne1cv) based on our evaluation of assessments in 10 units at 7 universities https://t.co/Kw0GvWiL0U by @SashaNikolicUOW et al. #engineeringeducation#eer
The @UNSWEngineering 🇦🇺 contingent is having a great time at #ASEE2023. It's so big 🤯 & we've so many cool ideas to follow up. Also fantastic to make connections between @UNSW Education Focused academics & US teaching focused faculty. @ASEE_DC@UNSWChemEng @unswpvcese #unswef
Great to share our findings on generative AI (ChatGPT) & assessments with colleagues from @Eng_IT_Sydney and @syd_health alongside coauthors Sarah Grundy & Marina Belkina. Read more at https://t.co/Kw0GvWiL0U. Thanks to Craig Campbell & @Tom_Goldfinch for hosting.
We also discuss how much prompt engineering is required to achieve something passable. There’s also a SWOT analysis about assessment integrity and integrating ChatGPT into courses, and discussion about how we need to adapt as a community of teachers and learners.
It was also fantastic to showcase the amazing engineering education practice and research being done at @UNSWEngineering. Our papers covered the breadth from student recruitment and experiences, to project-based learning, to teaching professional and technical skills.
Great time with colleagues from @UNSWEngineering and around the world at the @AAEE_HQ Annual Conference. Cool to also visit the nice UNSW Parramatta space. 1/2
#AAEE2022#EngineeringEducation