Student voice needs somewhere to go

This week, Dr Stuart Grey asks what happens after universities collect student feedback.

Drawing on new research into student-centred culture and longitudinal evidence about student engagement, Stuart explains why feedback needs a clear route to the people who can act-and why comments provide context that scores alone cannot.

In This Episode

Why collecting student feedback is only the beginning.
How evidence can become trapped at a different level from the decision it needs.
Why closing the feedback loop does not mean agreeing to every request.
What longitudinal research tells us about engagement scores and student change.
How open comments help teams distinguish between different barriers to participation.
Why bot protection improves one part of survey integrity, but does not establish representativeness.
Why LLM-generated survey comments make response-quality checks increasingly important.

Research

Across the Sector

Practical Takeaway

Take one recurring issue from last year's feedback and trace it from collection to decision and response. Give it a named route to the person or group that can act, carry enough context with the evidence, and show students what happened afterwards.

About This Recording

This episode was recorded by Dr Stuart Grey. The transcript was prepared from the final recording and lightly corrected for author names, research terminology, and readability.

Subscribe to The Student Voice Weekly.

Transcript

Hi, and welcome to episode 27 of Student Voice Weekly. I'm Dr Stuart Grey, founder of Student Voice AI, and this week I'd like to talk about where student voice actually goes.

So universities collect a huge amount of feedback. There are module evaluations, programme surveys, the NSS, student representatives, focus groups, complaints, and all the conversations that happen around teaching too.

But collecting is the easy part of this process.

The more difficult question is whether the evidence reaches somebody who has the authority, the resources, and the time to do something about it.

So a paper I read this week by Petra Nilsson Lindström, Åsa Bringsén, Maria Melén and Ann Ooms describes a student-centred culture as something that has to work across the university. The researchers worked with educators and first-year students, and then checked their ideas with more than seventy students.

The exact model is less important here than the central point: that students experience a university as one cohesive place, even though the university itself is made up of lots of different structures.

A student might raise an issue with a lecturer, but the cause sits in the programme timetable. A programme team might see the same issue repeatedly, but the solution requires faculty resources. Or a central team might identify a university-wide pattern, but the actual change has to happen locally within a department or school.

So when feedback stops at the level where it's collected, student voice can become a very well-organised record of things nobody has fixed, and we see that a lot.

So I still teach part-time at the University of Glasgow, and I think this matters in the day-to-day reality of teaching. Students do not always know which part of the university owns a given decision and, to be honest, staff don't always know either.

So if a student tells me that two deadlines are landing together, I can listen and I can explain, but I may not control the assessment pattern across the whole programme. The useful response is not simply to record that the student was heard. It is to make sure the issue reaches a group that can see the wider pattern and change it if necessary.

So that sounds obvious, but lots of feedback systems are designed around collection rather than movement of information.

So we ask a question, we receive a score, we produce a report, we might even make a dashboard. But none of those things tells us whether the evidence arrived in the right conversation with the right people.

The key thing is to give each recurring issue somewhere to go, a destination for this student voice information.

That means being really clear about ownership of the data. Who needs to see this? Who can make decisions based on it? What other evidence do they need? And how will the response get back to students?

The last question is important. If an issue cannot be changed, students still deserve an answer and an explanation. Sometimes there is a genuine constraint. Sometimes two groups of students want different things. Sometimes a suggestion would create a problem somewhere else.

That's totally fine, but the students need to know.

So closing the feedback loop does not mean agreeing to every request. It means showing the evidence was considered, what decision was made based on the evidence, and why.

There was a second paper this week that I found really interesting, by Chun Cao, Takacs Tamas and Ronghua Liang. That adds a useful warning about the numbers we use along the way when we're collecting the student voice.

The researchers followed nearly two thousand first-year students across three survey waves. When they compared different students, engagement and commitment seemed to move together. Students who were more engaged also tended to be more committed. Again, it makes sense.

But when they looked at change within the same students over time, that relationship disappeared.

So that distinction is really useful, I think.

Imagine that one group of students reports high engagement and a strong intention to continue with the course. It's tempting to conclude that increasing engagement will automatically increase commitment. But a pattern between students does not necessarily tell you what will happen when one student's experience changes.

So this paper does not make the survey result wrong. It changes the question the results can answer, right? We've got to be really careful on this distinction.

The risk is that we take a relationship in a dashboard or a spreadsheet and turn it into a causal story. Then we build an intervention around that story without checking whether it fits what students actually experienced.

For university teams, there are a few practical implications they could put in place.

They could try and keep measures stable when you want to understand change over time, separate different aspects of engagement rather than turn them into one general score, and put a concise open question beside the measure so students can tell you what affected their participation.

Obviously, I'm biased. I think this is really important.

That comment might point to workload, confidence, relationships, travel, paid work, the way a class is taught, or something else altogether.

The score shows you that something moved. The comments help you understand what is behind the movement, right? The underlying cause. That's where causality comes from here.

This is where I think student voice becomes much more than survey reporting.

If I were looking at the comments about engagement, I would not want one enormous heading called engagement. That's too broad to be useful. We have to try and break that down.

I would want to know what students were actually talking about and, particularly, the kind of response the comment requires. So what are we going to do about it?

A student who cannot attend because of timetable changes at short notice needs a different response from someone who does not feel confident speaking up in a seminar. A student struggling to access learning materials is describing something very different again.

And the comments tell you this, and we can categorise them at this level, which I think can be really useful.

Those comments may all sit beside the same engagement score, the question score, but they need to reach different people. So they should be grouped differently, and that's what our tools do.

This is why categorisation should help people find and read the relevant evidence. It should not flatten every comment into a label and pretend the work is finished. These labels are a guide to get people reading the verbatim comments.

So the data isn't the finish line. The reason we analyse the comments is to start better conversations and to get the right information to the right people so they can act on what the students are actually saying.

So there is a related development from Jisc this week around survey integrity. Online Surveys now has optional bot protection designed in to reduce automated submissions. This is a problem we're seeing more and more.

This is a sensible addition, particularly for public surveys where there's a credible risk of spam. But it also illustrates a wider point: that clean-looking data is not automatically trustworthy data.

Again, we've all seen, I'm sure, LLM-generated comments and things like that already in our surveys.

So a technical control can reduce one source of bad responses, but teams still need to look at who took part, who did not, whether the protection created some access barrier, and whether the response pattern makes sense.

So if I was running these Jisc surveys, I'd pilot it, test it on devices and networks students actually use, watch what happens to completion rates, and use it where the risk justifies this extra step, this extra technical control.

Because survey integrity has two sides. We want confidence that responses came from real participants, of course, but we also want genuine students to be able to take part without any unnecessary difficulty.

So that brings us right away back to our main point. A good student voice system needs more than a collection tool.

It needs a route from evidence to action. It needs enough context to interpret the numbers properly. It needs comments that remain connected to the themes drawn from them. And it needs a visible route back to students afterwards.

One practical thing to try at the start of this academic year is to take one recurring issue from last year's feedback and follow it all the way through the system.

Where was this first raised? Who received it? Where was it discussed? Who had the authority to decide on what was going on? What happened next? And what did students hear back?

You'll probably find that the weak point is not the survey question. It is one of those data handovers along the way.

So the takeaway this week is simple. Give every important student voice issue a named route so the person or group can act on it.

So that's it for this week. Thank you for listening.

All the papers, the Jisc update and all the written summaries are in Student Voice Weekly, the newsletter. You can subscribe at studentvoice.ai if you'd like them in your inbox each Friday.

And if you found this episode useful, please like and share wherever you listen to this, and send it on to somebody working in student experience, quality, surveys or teaching.

Thanks.