Earlier this year, we published a ‘StREAM in Conversation’ blog exploring the data trends for higher education. I had the privilege of chatting with a range of colleagues from across our client base as well as with the CEO of Kortext, James Gray.

Earlier this month, I had the further privilege of speaking with Professor Janice Kay CBE, special advisor to the Vice Chancellor at the University of Exeter, Director of Higher Futures and Advisor to the Kortext Academic Advisory Board. Janice is also a TEF Panel member. I wanted to pick up on the themes from the initial blog and think about the implications of those themes for higher education with Janice as well as draw on recent thought leadership insights in these areas. How should/could the sector respond? And what are the implications for Kortext?

Universities, and the sector as a whole, need to take the lead with generative AI

There’s no doubt about it – generative AI and large language models (LLMs) are here to stay. Their arrival on the HE scene has real implications for learning, teaching and assessment. The latter in particular is most problematic with the issues succinctly captured in Rod Bristow’s blog for HEPI on how the apparent threat being posted to HE by generative AI should be used to catalyse transformational innovation. Bristow argues that the sector needs to go further than the recommendation from HEPI that the Department for Education commission a report into how assessment will be affected by AI, arguing that university governing bodies act quickly to protect the currency of their degrees – if they haven’t already done so.

Kay agrees. Without a full-scale review of assessment practice, further grade inflation may follow, with the looming spectre of an Office for Students (OfS) review under Condition B4. This Condition requires universities to develop a regulatory framework that ensures that ‘relevant awards are credible’ and that that they are ‘credible at the point of being granted and when compared to those granted previously’.

However, for Kay, this is also an unparalleled opportunity for a revolution in learning and assessment – if universities embrace generative AI and LLMs and take a lead on how to effectively use the affordances from these technologies within learning, teaching and assessment. She is concerned that colleagues don’t sit ‘Canute-like’ hoping to push back the tide of generative AI. Rather, she believes that it is important to investigate ways to draw the best out of our learners, develop the requisite graduate employability skills and use AI to ask the right questions.

Doing this effectively will necessitate the provision of time and investment by senior university leaders and university staff alike – Kay’s ‘toppest tip’ for those seeking to address head-on the pedagogic challenges wrought by AI. Getting to grips with the technology and LLMs themselves is needed for academics to discover the core understanding that will drive pedagogic change and situate AI effectively within learning, teaching and assessment. (This Masterclass on the Explosive Growth of AI from Professor Jason Tangen at the University of Queensland is a good starting point for those keen to know more.)

This conversation – or at least the part about next steps – doesn’t feel dissimilar to conversations I was having in around 2014-15 at a UK university moving to small group learning and teaching. The challenge posed by the ready availability of content led to a major rethinking of the role of the academic – a pedagogic shift that focused not on the content, but on what students do with that content – how they apply it in context and how they develop graduate level skills. These discussions often resulted in conceptual questions about the nature of what it means to be an academic. Similarly, the ability of AI not just to write student assignments, but also to reflect upon, evaluate and incorporate feedback into the final version requires development of a different skill set – for staff and for students. How can they assimilate freely available content in ways that are novel and authentic, leading to deeper learning?

Part of the answer, according to Kay, could be to make the technology map used to arrive at the final assessment submission part of what is assessed. Making the process itself more transparent would enable assessment of the skills used to arrive at the submission, review of the questions asked of the technology, consideration of the refinements and exploration of how the assessment in its entirety has deepened the student’s learning and how they arrived there.

Learning is fundamentally social

Having been part of the TEF assessor panel, Kay has a lot of insight and experience into the state of play across the sector. In this post-pandemic world those Universities able to provide a ‘judicious mix’ of online and face-to-face learning opportunities are the ones who appear to be embracing the positive lessons from the pivot to online learning from the pandemic and then the more proactive teaching approaches that emerged as campuses opened again.

There’s something here about recognizing that learning is fundamentally social – ensuring that the opportunities that exist for students to be together with their tutors and then making best use of that togetherness through application of knowledge and collaborative learning, all of which help build belonging. Digital technologies have a key enabling role here as we know – scaffolding learning, collaboratively engaging – with content, with peers and with tutors, helping students transition through those liminal spaces to where their learning is secure and then testing that learning in ways that allow metacognitive learning and skills development to be evidenced, rather than a student’s ability to recall a set of facts.

Data can help

Against this backdrop is data in its many forms. Data can be incredibly useful – if the right data is collected and if there is a clear understanding of how it should and could be used. Having too much data can be as problematic as not having enough. Knowing what data is needed – and, critically, what questions you want to ask of that data – is when it starts to become possible to leverage and realise the data’s inherent value. The key with data is to turn it into information and knowledge that provides a set of insights which can provide the wisdom to use the data in such a way that it begins to impact those factors that you want to address. For example, analytics within the Kortext platform make it easy for academics to check if students are engaging with their learning and to spot common issues where students may need further guidance.

In this context, continuation and completion measures remain key data points for higher education providers and for the government. Identifying which students would most benefit from additional outreach and support activity at any given moment and then working with them to effectively access and implement that support is where learning analytics platforms, such as StREAM by Kortext, can help. A recent randomized control trial conducted by TASO identified that learning analytics can effectively identify students at risk of withdrawing from their studies or underperforming. The study then sought to evaluate the impact of the consequent intervention on engagement and success. TASO concluded that further research is required in terms of the post-identification interventions that follow.

Caroline Reid, Associate Dean in the Faculty of Health and Social Sciences at the University of Bedfordshire, recognises the importance of lead data indicators that you see in StREAM in the light of these sector-imperatives: arguing that the data you see in today StREAM is next year’s continuation data. So universities who want to meaningfully turn the dial on their continuation metrics, need to use data effectively to understand how best to garner those improvements.

Partnering for Success

Like Reid, Kay is a strong believer in partnering with external providers like StREAM by Kortext, when it comes to effectively using data to identify students at risk and then join the dots around student support to ensure that barriers to participation are effectively overcome without any sense of a deficit model coming into play. In her opinion, there is little point in a university reinventing what StREAM already does effectively, often in less efficient and more resource-intensive ways. But ensuring that senior leaders and budget holders understand why and how StREAM can help is essential to their buy-in, particularly when many universities are so strapped for cash. Recognising that focus on continuation and completion is here to stay – in Kay’s words, ‘it is likely to remain in TEF 2027’ – requires a long-term perspective and a long-term investment in analytics.

The importance of TEF as a driver in these discussions cannot be overstated. Kay observes that one feature evident in the 2023 TEF submissions for those institutions awarded gold, was the narrative around and approach to student support, typified in the TEF panel’s summary statement of why Teesside University were awarded triple gold. Teesside University have partnered with us to deploy StREAM as their analytics system and the TEF panel cited their use of ‘a learner analytics system to make informed improvements’, as one of the reasons why their learning environment and academic support was considered by the panel to be ‘outstanding’.

Engagement is at the heart of the student experience

Kay is keen to promote how TEF can act as a catalyst for change through how universities use their data. She argues that student engagement is at the heart of the student experience and student learning. Understanding what impacts student engagement – either positively or negatively – in effective ways is thus the pivot for learning. Effective means clearly identifying those students who are only marginally engaged as soon as possible and putting rectifying measures in place. It also means ensuring that learning analytics are deployed consistently across the Institution’s subject base. TEF Indicators can spotlight those subject areas that are significantly materially below a subject benchmark for learning environment and academic support. Jo Midgley, Registrar and PVC (Student Experience) at UWE Bristol, a StREAM client, explains how StREAM has helped her university meet this imperative:

We have been able to provide more specialist support for those who might not identify themselves as needing it, and support others to persist and achieve more than they might have without an intervention.

Global in Learning

Finally, I asked Professor Kay what she saw ahead for HE policy in what is likely to be an interesting election year. Like my previous contributors, she is aware that the primary focus for all political parties is unlikely to be HE. But that shouldn’t, in her opinion, stop universities from seeking to drive the conversations and the sector forward. Not being top of the party manifesto doesn’t mean universities lack power. There are other ways in which universities can come together to influence and lobby governments, to highlight the ways in which their work is boosting the economy, to shine a light on negative impacts of current immigration policy on UKHE’s position in the global HE market and on universities’ financial stability. Above all else, she argues, we need to be ‘global in learning’:

Even if HE is not front and centre [on the policy agenda] what is our collective position as a sector and how can we lobby to show the value of international students and our presence on the global HE stage?

Following the recent unification of FE and HE in Wales under the Commission for Tertiary Education and Research (due to become operational in August 2024) and similar conversations in Scotland, Kay would certainly be a proponent for further exploration of this approach in England. She asks whether some of the challenges around funding post-16 education might be addressed by considering current structural funding of HE, FE and HE in FE. In a sector where there may be no (more) money perhaps some radical thinking really is required.

Conclusion

Working thoughtfully in and creating productive partnerships with education technology partners like Kortext, where technology-related decisions are firmly rooted in pedagogy, will be needed to drive forward digital transformation within our higher education institutions. Understanding what universities want to achieve and then working collaboratively to support the accomplishment of those aims can help ensure that the development of educational technologies are both fit-for-purpose, but also future-proofed.

Now that StREAM is part of the Kortext group, there is a significant and timely opportunity to bring data around both learner engagement and learning analytics together. Understanding more deeply how students are engaging with their learning – what they are reading and what learning activities they are engaging with, exploring how time-on-task is impacting assessment outcomes, and surfacing all these insights in ways that are meaningful for a range of staff with different responsibilities within a university has the potential to deliver a richness of insight that can inform curriculum design and development as well as help coalesce relevant activity around individual student success – and measure impact.

How we can help

Find out more about the StREAM student engagement analytics platform can enable your university to identify students at risk of under-performing academically and support them to harness those student support functions that will be key to their individual success.

See StREAM in action – book a demo!

Find out how Kortext’s cutting-edge, AI-powered study tools can transform learning and teaching at your institution.