Last October, I had the honour to give a keynote address at ALTC25. As my keynote will be radio-streamed again at ALTC25 Revisited on 25 February, I decided to revisit the talk and share a summary of it here for anyone who is interested. You can watch the keynote on ALT’s YouTube channel (link at the end of this post), and the slides are available here.

The TLDR:
In my talk, I share stories from refugee-background learners, migrant teachers and my own journey towards becoming a learning designer in a global higher education (HE) context, to make the case for designing from the margins, not the centre. Drawing on the Capability Approach, I argue that behavioural, emotional, social and cognitive engagement all fuel each other – and that neglecting any one of them is risky, especially for students living with high “lifeloads”. I suggest three practical moves for more socially just and inclusive online learning: finding a kinder balance between flexibility and structure, embedding “warm” human support into course design and delivery, and intentionally creating accountable spaces where students can matter, be recognised and express themselves freely. A more detailed commentary on the talk follows.
Starting from the margins
What does engaging online learning mean when we consider students living in precarious, crisis‑affected contexts? The photo above, taken in a classroom in a Migrant Learning Centre in Mae Sot on the Thai–Myanmar border, sets the scene: a well‑thumbed, Pacific‑centred paper wall map sits alongside an equally well‑used computer. That map prompts us to ask who is really “on the margins” here; in a setting many of us in well‑resourced HE institutions in the Global North might label “peripheral,” this classroom is in fact the hub. For me the image is a quiet provocation: it invites us to feel, even briefly, what it is like when we are the ones at the edge of someone else’s map – and then to ask what that might mean for how we think about inclusion, centre and margin in our own designs.
My starting point, from many years of experience as a learning designer, is that: (a) online learning design is usually good design for all modes, and (b) designing for the most challenged learners tends to improve the learning experience for everyone. ALT members will know that there is plenty of evidence for both of these notions, and that the latter is one of the fundamental principles of universal design for learning (UDL).
I then situate myself in the story. I grew up as a white child in apartheid South Africa, thinking I had a pretty ordinary childhood, but over time realised the unjustified privilege built into that “ordinary” life. An exchange year in Switzerland was my first major wake‑up call, followed by adult literacy work in the NGO sector back in South Africa, which aimed to support adults who had been denied a decent education through the apartheid regime’s iniquitous Bantu Education Act. Later, my own distance learning experience turned me into an online learning enthusiast and led me to wonder about the possibilities for online education as an instrument for widening participation.
That personal arc led me into the area of research I have been engaged in for the last ten years: trying to make sense of the barriers and enablers to online education for people in contexts of displacement and protracted crisis, and seeking ways to widen engagement in online HE/ professional development for people in such settings. My PhD research (see thesis and earlier blog posts) investigated student engagement on an online Master’s programme run by the University of Leicester; my research participants there were displaced people holding Sanctuary Scholarships and studying from around the world.
Subsequently, I joined UCL as a research fellow. Our small project team works in partnership with an NGO, the Inclusive Education Foundation, on the Thai–Myanmar border, researching and developing Co-designed Massive Open Online Collaborations (CoMOOCs) aimed at teacher professional development in displacement settings. The research is part of the RELIEF 2 project funded by the ESRC, and also has funding from the FCDO via the ERICC programme. (See our blog, CoMOOCs for Transformative Professional Development for more about that project.)
Taken together, these experiences have decisively shaped how I think about online engagement – and in the rest of this post, I share some of the key ideas and stories that I brought into the ALTC25 keynote.
Four kinds of engagement
The talk was framed around a four‑dimensional model of engagement – behavioural, emotional, social/collaborative, and cognitive – adapted from Redmond et al. (2018). A quick poll with the audience confirmed my expectations that HE tends to reward behavioural engagement most (turning up, submitting assignments, doing the “right” things), with cognitive next, and social and emotional trailing behind. The practice thus runs counter to a growing body of literature (including my recent research), which provides strong evidence that engagement in one dimension fuels engagement in the others, and disengagement in any one dimension can fuel disengagement across the board.
This is the thread running through the talk: if we neglect emotional or social engagement, we shouldn’t be surprised if behavioural and cognitive engagement start to unravel, especially for students on the margins. To understand this from a social justice perspective, I draw on the Capability Approach from Amartya Sen and Martha Nussbaum, and Melanie Walker’s (2006) seminal work in applying capabilities to the HE context.
Capabilities, not just skills
According to Sen and Nussbaum, capabilities are not just personal skills and abilities that can be developed and practised: they are the real freedoms that are socially, politically and economically enabled. These freedoms enable people to achieve valuable “functionings”, such as being educated, healthy, or able to participate in society. The distinction matters because a student may have the ability to study without having the freedom to do so, thanks to political, economic, or social constraints. My example from apartheid South Africa, of people who physically could swim but were legally barred from public beaches and pools, makes this tangible.
I therefore ask: If we understand capabilities as both personal abilities and socially enabled freedoms, what capabilities are needed for online engagement in each of the four dimensions? We must acknowledge that universities can’t fix problems arising out of structural issues such as poverty, war, and state policy – but we can recognise that our learning design can either mitigate or compound those external constraints.
Behavioural engagement: resilience and lifeload
I argue that behavioural engagement is underpinned by the capability of “educational resilience” (Walker, 2006): the capability to navigate study, work, and life; negotiate risk; persevere academically; and stay responsive to opportunities while adapting to constraints. Lifeload – the total pressure of everything in a student’s life, with university just one piece of that puzzle (Kahu, 2013) – becomes the central idea here. Covid‑era research shows that, when students were all learning online, they consistently prioritised lifeload over learning load, which confirms what many of us have observed anecdotally.
The stories from my PhD research make this visceral. There was Julian, studying from a refugee camp in Malawi, without electricity or reliable internet, studying on a borrowed laptop. He successfully completed the four taught modules on the online MA, but when time started running out to complete his dissertation, he explained that “situations complicate me. Psychologically I am failing.” There was Sami in Malaysia, who had founded a school for refugee children. He worried about paying his rent while trying to read academic papers, describing the dizziness he felt when stress overwhelmed his concentration. There was Zain, a Syrian refugee in Berlin, who was homeless for a while during winter because couldn’t get a flat. (“No-one wants to rent a flat to someone with a beard like me.”) He used to go to a freezing train station to access the free wi-fi and download his course materials. Some of these students exited early with a certificate or diploma; most graduated with a full Master’s, but none of their pathways were straightforward.
Turning to the Thai–Myanmar border context, I share the astonishing story of a whole class of students “sneaking” out of a refugee camp, then walking miles through mountainous areas and forests to access electricity and wi‑fi in a nearby village, risking trouble with Thai authorities simply to engage in an online course. That is contrasted with the practical example of a collaboration with Arizona State University to pilot the use of the Beekee Hub/ Sunspot – a portable offline Moodle server – which allows people in a locally connected classroom to access course materials without wider internet infrastructure.
These stories illustrate the profound desire of so many people in settings of displacement to learn online, alongside some of the often formidable barriers to engagement. My design recommendation here is to increase behavioural engagement is to find the kindest balance between flexibility and structure, for example by offering more flexible deadlines and clearer exit points (e.g., certificates, diplomas) without sacrificing the cohort‑based structure that supports social learning.
Emotional engagement: mattering and “warm support”
For emotional engagement, I focus on the capability of “emotional health”: students being able to experience emotions that support learning rather than being dominated by anxiety, fear, or trauma. The stories above illustrating behavioural engagement are already saturated with emotional struggle; the two are inseparable. This links to a body of literature about students’ need not just to belong but to matter (e.g., Austen et al., 2021) and the research suggesting that trauma is far more widespread among online learners than we often assume (e.g., Reinhardt, 2022).The call from bell hooks for a pedagogy that respects and cares for students’ souls as a precondition for deep learning is fitting here.
My main design recommendation for enabling emotional engagement is to embed “warm support” into course design and delivery: this is the kind of support that sits between informal “hot” support (such as gossip, or WhatsApp groups) and “cold” support (such as formal institutional policies and programme handbooks) (Baker et al., 2022). Warm support extends to the tone and texture of the whole learning design, including clear navigation structures, the use of plain language, and prioritising the human voice. For example, in an experiment at Leicester University, when a programme team re-created the entire dissertation handbook as a series of short podcasts, this dramatically reduced student queries: students said that while they struggled to read the handbook, they were highly motivated to engage with the information that was presented in a “warm” and “human” manner through the voices of their teachers and other experts in the field.
Social engagement: affiliation, recognition, and accountable spaces
For social and collaborative engagement, I suggest that the key capability is, in Nussbaum’s words, “affiliation and recognition”: the ability to interact with others to learn and solve problems, while being treated with dignity and entering into relationships of mutual respect and trust. I argue that, although universities can’t control what happens in students’ wider lives, we can shape how affiliation and recognition are experienced within our courses, even online.
The voices from my research into student experiences on the online Sanctuary Scholarship programme again shed light on this point. Nadia’s description of discussion forums as “very, very helpful” because they put her with “other human beings” is exactly what any online educator would hope for. Maryam’s journey is more complicated, however: she initially dismisses forums as a waste of time because they don’t count towards the assessment, then gets excited when she reads about her classmates’ backgrounds on the forum, but then is suddenly reluctant to introduce herself, feeling that “I am nothing in comparison” to them. This is an unsurprising response, considering the well‑known social media dynamic of a heightened awareness of inequalities and different capabilities in online social settings. I found that even a simple icebreaker like “Describe the view from your window” could trigger a cascade of doubt and difficult emotions for some students.
Some people promote “safe spaces” to mitigate these barriers, but, as Elise Ahenkorah argues, truly safe spaces are probably impossible to create. I would go further and suggest that such spaces may even be ill‑suited to genuinely transformative higher education. “Brave spaces” are often proposed as an alternative, yet this notion tends to place the burden on those experiencing trauma to be “brave”, thereby failing to address the underlying problem. Following Ahenkorah, I instead advocate for “accountable spaces” (which I have discussed elsewhere in my blog and in a presentation to the Knowledge Exchange Network). These are collaborative spaces built on shared principles for working together, where facilitators explicitly acknowledge the fragility and complexity of the humans in the group, and where respect and recognition are actively modelled by staff. Well‑established approaches to collaborative work – such as deliberative democracy practices, World Café discussions, and “lean coffee” sessions – offer practical ways of beginning to enact accountable spaces in online education. The co‑design work on the Thai–Myanmar border illustrates how curriculum design itself can become a site of mutual learning, affiliation, and recognition.
Thus, my recommendation for increasing social engagement is to build accountable spaces in which students can feel both safe and answerable to one another as an online learning community, while they critically co‑construct knowledge throughout their courses.
Cognitive engagement: knowledge, imagination, and social justice
Finally, I argue that cognitive engagement is underpinned by Nussbaum’s capability of “knowledge and imagination”: the freedom to think, imagine, and produce academic and professional work that matters to oneself and to others, and to be an active inquirer without fear of reprisal or censorship. Again, the social thread is explicit in the formulation of the capability, and rightly so: none of us working in HE want to produce graduates who only gain something for themselves; we all want HE to equip people to better contribute to wider society.
In relation to cognitive engagement, I share in the joy of students like Kareem, who feels that the way he debates and argues “becomes part of your DNA,” and Mohsin, who says he simply loves the fact that he is “free to think” on his online Master’s programme. But I also surface the quieter tension in Nadia’s story: her careful skill in “knowing both sides” so she can tailor what she says depending on the audience, which was a survival strategy she had honed over her lifetime having grown up in an authoritarian state, but belied an undertone of self‑censorship in the online HE setting. That brings me back to accountable spaces: if students don’t trust the space, real critical inquiry is risky and likely to be avoided.
My closing argument is deliberately circular: the three recommendations I made for the other dimensions – finding a kind balance between flexibility and structure, embedding warm support, and building accountable spaces – are also my recommendations for cognitive engagement. In my view, universities are already very good at developing students’ cognitive skills – in the sense of personal skills and abilities that can be developed and practised (capabilities in the everyday sense of the word). What is missing is deliberate attention to the real freedoms that are socially, politically and economically enabled – the social‑justice‑infused conditions that make deep cognitive engagement possible.
In conclusion, I suggest that when we design online learning that supports students’ capabilities not only for behavioural and cognitive engagement, but also for emotional and social engagement – especially for students at the margins – we are not just making better courses. We are modelling social justice in everyday practice, and in doing so, we are quietly expanding the possibilities for greater social justice in higher education and beyond.
If you’d like to watch the recording of the original presentation – here is the link to view it on YouTube.
Many thanks to ALT for inviting me to present this keynote, and thanks to everyone who has taken the time to read this. Please feel free to post comments or questions below.









