Belonging is not a single number

Audio briefing based on Student Voice Weekly issue #8.

This Week

This week, the episode discusses belonging, mature student induction, DMU block teaching, Jisc DEI. 66 studies show retention work needs richer belonging evidence The main topics are grouped below by student voice practice, research, sector developments, archive context, and practical application.

Main Topics Discussed

Student Voice Practice

I spent the first half of this week at the Future Facing Learning and AI in Higher Education conference at Teesside University, where I delivered a keynote on who will control the AI infrastructure of higher education.

Research Spotlight

Sector Watch

From the Archive

Practical Application

Universities often ask whether student comments are truly secure when processed by AI.

Subscribe to The Student Voice Weekly: https://www.studentvoice.ai/blog/newsletter/

Transcript

Hi, and welcome to Student Voice Weekly. I'm Dr Stuart Grey, founder of Student Voice, and this week I want to talk about belonging: why a single belonging score will not predict who leaves.

Today I'd like to talk about a trap I see universities fall into. We know belonging matters, and we know it links to continuation. So we add one question to a survey, track the score, and hope it works like an early warning system.

The key thing is this: belonging is not one thing. It changes over time, and it is not evenly distributed across groups. So if you reduce it to a single average score, you might get a comforting trend line, but you lose the detail you need to act.

The main research story this week is a systematic review across 66 empirical studies on belonging and retention. The headline finding is that the evidence base is conceptually scattered. Different definitions, different measures, different theories for how belonging connects to staying or leaving.

That might sound academic, but it has a practical consequence. When we treat one belonging item as strong evidence, we are often treating a shaky proxy as if it is a diagnostic.

So what should UK universities do with that?

First, be clear what job you are asking belonging data to do. In practice there are three.

One is descriptive: are students generally feeling connected and accepted. Second is evaluative: did a change to teaching, support, or induction improve things. Third is predictive: can we identify who is likely to disengage and withdraw.

A single score can help with the first job as a rough pulse check. It is much weaker for evaluation unless you know what changed for whom. And it is weakest as prediction, because withdrawal is usually driven by a specific friction that becomes cumulative.

This is where listening to what students are actually saying matters. Students rarely frame it as, my belonging is low therefore I will leave. They describe the mechanism. I cannot make group meetings because of my commute. I do not understand what a good answer looks like because feedback is vague. I asked for help and got passed between teams, so I stopped trying.

Those are belonging stories, but they are also process stories. Teaching design, assessment clarity, staff responsiveness, timetabling, admin, digital access. If you only watch the belonging dial, you miss the thing you could fix.

There is also a timing problem. A lot of institutions measure belonging once a semester, or once a year, and treat it like a stable trait. But belonging is spiky. It drops at transitions, around the first assessment, when expectations are unclear, when students experience something they see as unfair, or when they feel invisible.

So make sure you measure it where risk actually shows up. Early in semester, at first assessment, when feedback lands, at key handovers between levels, and when students are making course choices. If you capture the story at those points, you can see what is changing and why.

The second research spotlight this week makes the same point through a specific group: mature students and induction. The paper is basically saying, yes, induction can build belonging. Pre entry activity helps. Staff relationships matter. Clear work on connection and expectations helps.

But even good programmes can still leave mature students feeling like outsiders, because institutional defaults are built around the school leaver, flexible timetable, always on campus model.

This is the core belonging lesson I want you to take away. Belonging is not just more welcome activity. It is whether the institution's defaults signal: you are normal here, or you are an exception we are trying to accommodate.

A simple example is session timings. If key induction activity assumes people can be on campus at a fixed time in the middle of the day, you are already excluding some carers, commuters, and students in work. When they do not attend, it gets misread as disengagement rather than structural mismatch.

So the practical implication is segmentation, but not segmentation as a reporting exercise. Segmentation as design. Ask what the transition looks like for different groups, and change the defaults where you can, rather than relying on bolt on support.

Now a sector signal that connects to this: De Montfort University has published its evaluation of block teaching. They are reporting improvements across survey categories and NSS themes, and improved continuation. They also highlight a 15 day feedback turnaround.

I am not here to say block teaching is the answer. The risk is copying a delivery model as if it is a magic switch. The interesting point is what it forces operationally. Tighter assessment and feedback cycles. Clearer prioritisation of what matters in a short period. Less cognitive overload from juggling multiple modules and competing expectations.

That matters for belonging because academic belonging is a big part of the story. Students often feel they do not belong when they cannot see how to succeed, or when they feel permanently behind.

Make sure you separate speed of feedback from usefulness. Students care about timeliness, but what they really care about is whether feedback helps them do the next thing. Fast but generic feedback does not build belonging. It just completes a process.

So if you are looking at block delivery, the question is not only timetable structure. It is whether you can deliver clarity and usable feedback at pace, consistently, without burning staff out. That is where the experience gains come from.

The other sector watch item this week is Jisc retiring Digital Experience Insights at the end of July. If you use it, you will need a plan.

Yes, export your data. But the bigger issue is that a piece of sector evidence infrastructure is disappearing. Institutions have built reporting rhythms and digital enhancement narratives around it.

Do not treat this as simply swapping one survey tool for another. Digital experience is tied to belonging and continuation because if students cannot reliably access systems, cannot navigate the VLE, or experience online delivery as excluding, that becomes part of the belonging mechanism.

So a sensible replacement strategy starts with a different question: what decisions did we use DEI to make, and what is the minimum evidence we need to keep making those decisions well.

Now, what do you do with comments and free text, the stuff that is messy but often more diagnostic than the score.

If I was analysing belonging comments, I would separate them into three types.

Connection: do I have peers, do I feel known by staff. Competence: do I understand what is expected, can I see how to succeed. Legitimacy: does the university feel like it was built for people like me, do the assumptions fit my life.

If you lump all of that under community or wellbeing, the action becomes vague. More socials, more signposting, more encouragement. Sometimes that helps. Often the real issue is operational.

Also watch for belonging as a signal of trust. When students say they do not feel they belong, sometimes they are saying they do not trust the institution to respond fairly when something goes wrong. That is why consistent communication, transparent assessment, and credible handling of problems matter. They build trust, and trust supports belonging.

So here is one practical thing to try this week.

Keep your existing belonging question for trend. Do not throw it away. But add two short open text prompts at one transition point you already measure.

Prompt one: what is currently making it easier for you to feel part of your course. Prompt two: what is currently getting in the way.

Then do a quick sorting exercise. Put the barriers into four buckets.

Timing: timetable, deadlines, availability. Clarity: expectations, briefs, feedback quality. Consistency: different rules across modules, mixed messages. Usability: systems access, admin processes, navigating support.

That gives you a short route from comment to action, without pretending belonging is a single construct you have perfectly measured.

And if you want to go one step further, segment just one group. Mature students is an obvious place to start, given the research. The point is to test whether your defaults are excluding people, and then decide what you can redesign.

That is it for this week. The full set of links and summaries is in Student Voice Weekly.