Getting the right student feedback to the right teams

Student Voice Weekly, episode 35, with Dr Stuart Grey.

Dr Stuart Grey explores how school and programme reports connect local survey comments with institution-wide analysis. He explains the role of organisational information, consistent categories and sentiment, reading the original comments, and bringing local interpretation back into institutional decisions.

Chapters

• 00:00 - Getting feedback to the right teams
• 00:43 - A head of school investigates organisation
• 01:35 - The work involved in separate local analyses
• 02:17 - Shared analysis and organisational information
• 03:57 - Reading the comments behind a category
• 05:03 - Programme comparisons and group sizes
• 05:47 - Connecting local and institutional responsibility
• 06:50 - Learning from positive findings
• 07:12 - Glasgow's student feedback loop
• 07:42 - Taking reports into decisions
• 08:04 - Further reading and next week

Features and further reading

• School and department feedback reports.
• How one business school redesigned its student feedback loop.
• Enhancing Student Voice: Developing the Student Feedback Loop, a practice-based case study. Alison Gibb, Gail Angus and Karen Clancey, Student Engagement in Higher Education Journal, 7(1), 79-87, September 2025. DOI: 10.66561/sehej.v7i1.1421.

The school example in the episode is illustrative. The research account describes practice and implementation challenges; its impact still required fuller review.

Subscribe to The Student Voice Weekly. Browse the podcast.

Transcript

Hello, and welcome to Student Voice Weekly. I'm Dr. Stuart Grey, founder of Student Voice AI, and today I'd like to talk about getting student feedback to the people who can use it. In particular how we move between the institution-wide analysis and the school or programme level reports that colleagues need for their own decisions. So I still teach part-time at the University of Glasgow, and the relationship between the different levels of the institution is something I still think really matters.

A university can have a sensible overall priority, and an individual programme team can still need a much more specific explanation of what the students are experiencing. So both things can be true. So let's take an example. So imagine you're a head of school preparing for a higher level education committee.

You've seen the institutional analysis and organisation and management, for instance, is one of the themes that stands out. You now want to understand what that means within your school. So are students describing changes to their timetable? Are they struggling to find assessment information? Is there a problem with how different teams communicate?

And are those experiences spread across the school or concentrated in a particular setting? So you can see how quickly a useful headline turns into a series of more detailed questions and requests for more information. So those questions are about the survey comments you've already collected. The data's already there.

The next step is to make that evidence available in the form colleagues at every different level of the organisation can investigate. So one approach is to ask each school to work through its own responses, its own comments. That brings local knowledge into the process, which is valuable, but it also means several teams repeating the categorisation work if they do it at all, using different definitions and producing very different reports that are very difficult or impossible to compare later.

Another approach is for institutional research teams to prepare every local report individually. They can preserve consistency although each new request adds more work. Then they can do another programme split, another theme, another set of comments to extract. The central team can end up spending a lot of its time producing these small views of the data.

So that's why we've built school, department, and programme reporting into Student Voice AI directly. The starting point is a shared analysis for comments. We then use the organisational information supplied with the data to create relevant local views, keeping the category framework and sentiment method entirely consistent across all of them.

That's really important. The organisational information also deserves a little attention because it determines what a local report actually means. So a student's home programme might sit in one school while a module is taught by another department. It's very common. Both are useful ways of looking at feedback, but they answer different questions.

So we have to build our outputs based on that information. So we'd begin by agreeing the reporting structure with the institutional research team. Which fields are available? What does each one represent? How should joint programmes be treated? Is it the home or the student's department? Has the organisation changed since the survey was collected?

Is there any information missing from this organisational structure, right? Have things been updated? So that sounds like preparation work, and it is. We spend a lot of time doing it, but it prevents a very practical problem later on when teams don't have access to the right information, and they're talking at cross purposes because they're talking about different sets of data.

So once that organisational structure is clear, we can then produce school or programme views from the same classified data set. The local team can see its themes, sentiment, and relevant comments, while the central team keeps a wider picture available to them. People have a common starting point for the conversation 'cause they know the same models, same approach that have been used for both, but they can see the data that's relevant to them.

So now let's go back to our hypothetical head of school. They open the local report and look at organisation. They read the comments. There are negative comments about timetable changes, but there are also positive accounts of clear assessment schedules. So that distinction is useful to them, suggests different processes, potentially involving different people are at play here.

So reading the comments is what makes a category more useful. So the team can see how students describe the changes, what consequences they mention, and where their experience varies. A comment about a late room change might describe a minor inconvenience for one student and a missed teaching session for another.

So that context is really important, and that's why we always want people to be reading the verbatim comments as well. The aim of all this is to examine these patterns across all the responses. You want to understand the range, including the comments that complicate your initial interpretation. So a striking quote might help communicate a finding, but you can over-index on that, and we need to keep all of the comments in mind and not just pick the ones that are most eye-catching. The available programme or year information can also help the team look more closely.

So perhaps the same concerns appear across several programmes. Perhaps one part of the school has different experiences. But we need enough comments to support those comparisons and we need to show the size of the groups alongside the findings. So a larger programme will often contribute more comments simply because more people responded.

That is why a raw count on its own isn't a very good measure for comparison. And if a small group has only a handful of responses, a precise looking percentage can give a misleading sense of certainty. So sometimes the more useful view is at a broader level. And in our outputs, we always show the number of comments the analysis is based on, so you can go up a level if need be.

There's also another benefit to keeping the local institutional analyses connected in that it gives a way to distinguish the level at which something can be changed. So a programme team may be able to improve the information it gives students about an assessment, but timetabling issues frequently are a central problem, and there's nothing the academics on the ground can do about that.

So if each school receives its report in isolation, that shared issue could be treated as several separate local problems with a common analysis. But with a common analysis, the institutional research team can bring those findings together and ask whether a coordinated response would make more sense at an institutional level.

Local colleagues also have information that the central analysis cannot supply by itself. They may know when a process changed, which students were affected, or why one set of comments describes a different experience. Their interpretation can improve the institutional account as well. So that's a useful exchange in both directions.

So that's why we always produce both. It's not one or the other. It's both for both sets of users. So I would give positive findings the same attention as negative findings. If students repeatedly describe a helpful approach in one area, read those accounts and explore what colleagues are doing and try to spread that best practice.

We still need to understand the context before assuming anything will transfer, but having some evidence is a good starting point, right? To start those conversations. A research example connects up well with this. So Alison Gibb, Gail Angus, and Karen Clancey describe changes to the student feedback loop at Glasgow's Adam Smith Business School, so very close to home.

Their practice account includes reporting at course, programme, subject, and school levels with revised processes for taking findings into decisions later on. They also describe difficulties following staffing changes and say the impact needed further review. So the useful example is how the reporting governance were organised.

So producing the report is one part of the work. Colleagues also need to know where the finding will be discussed, who can investigate it, and how a decision will come back later on into the wider process. The detail you share should be fit for purpose, so the right people should receive the right information.

That's basically all we're going for. So you can find out more about how our school and department reports work on the Student Voice AI website. I've also included a link in the show notes along with the research account we discussed from the Adam Smith Business School. You can also sign up for the weekly newsletter on the website.

Next week I'll look at how sector comparisons can add context to that local picture and what we need to check before interpreting a benchmark. So thanks for listening to Student Voice Weekly. I'm Stuart Grey, and I'll see you next week.