What ethnic-minority students mean by belonging, and what surveys often miss
Updated Apr 11, 2026
Universities cannot improve belonging well if they measure it too narrowly. At Student Voice AI, we see the real problem surface when students explain why a course, campus, or peer group feels open to them or closed off. A 2025 paper in Studies in Higher Education by Jente De Coninck, Peter Stevens and Wendelien Vantieghem, "Sense of belonging defined: how ethnic-minority students conceptualise belonging in the university", starts in the right place: not with the survey instrument, but with students' own definitions. For UK universities using student voice data to improve inclusion, that shift matters because it makes belonging easier to measure well, diagnose accurately, and improve in practice.
Context and research question
Belonging is now a common term in higher education strategy, access work, and student experience dashboards. The problem is that it is often measured before it is clearly defined. Institutions may ask one or two survey questions about whether students feel part of their university, then act as though those answers fully capture the construct. The risk is obvious: if the measure is thin, the response will be thin too.
De Coninck and colleagues tackle that gap directly. Working in Flemish Belgium, where ethnic-cultural-minority students continue to experience weaker academic outcomes despite an open-access higher education system, they ask how ethnic-minority students actually conceptualise sense of belonging, and what that means for more valid measurement. The study uses qualitative content analysis of 69 interviews with ethnic-cultural-minority students. The context is not the UK, but the practical problem is familiar, especially given recent evidence on how ethnic-minority students build belonging over time. If universities want to improve belonging, they first need to know what students themselves are counting as belonging. That gives teams a stronger basis for designing questions, interpreting results, and choosing interventions that match lived experience.
Key findings
The first finding is conceptual, but it has direct operational value. Belonging is not a single feeling that can be reduced to one neat survey item. Students described it across social, personal, and academic domains, rather than as one undifferentiated sense of fit. That matters because a single headline score can hide which part of the student experience is actually under strain.
"social domain is the most important, followed by the personal domain and to a lesser extent the academic domain."
That ordering matters. The paper suggests belonging is driven first by the quality of students' social world, not only by their formal academic role. If the social domain carries most weight, universities cannot assume that belonging will improve simply because teaching is well organised or support services exist on paper. Peer relationships, everyday recognition, and whether students feel at ease in ordinary campus interactions are likely to be doing a great deal of the work. For institutions, the benefit of seeing that clearly is better prioritisation: social design, not just formal provision, becomes part of the response.
The personal domain coming next is also significant. Belonging is partly about whether students can be themselves without excessive self-monitoring. That has immediate relevance for UK institutions thinking about ethnic-minority attainment gaps, inclusion, and campus climate. Students may be attending, passing, and engaging outwardly while still feeling that belonging depends on managing identity, code-switching, or deciding when it is safe to speak, a pattern that also appears in conditional belonging among minority ethnic STEM students. That is useful for practitioners because it moves belonging beyond simple participation metrics and towards the everyday conditions that make participation sustainable.
The academic domain still matters, but the study places it third rather than first. That is a useful corrective for institutions whose belonging measures focus mainly on teaching, staff contact, or academic integration. Survey instruments can look precise while still under-measuring what students actually mean. The authors therefore argue for a broader conceptual model of belonging, one that can capture more universal aspects of the construct across groups and contexts. For survey teams, the practical implication is straightforward: a narrower measure may be easier to field, but it is harder to trust.
For student voice work, this is the most practical takeaway. A scale can tell you that one cohort reports weaker belonging than another. Open comments tell you which domain is breaking down. They reveal whether the issue sits in friendship networks, identity safety, recognition, support, or academic connection. That is where systematic comment analysis becomes valuable, because it helps teams move from reporting a problem to understanding what to fix first.
Practical implications
First, UK universities should audit their belonging questions against the three domains identified here. If a pulse survey asks only whether students feel part of their course, it is probably capturing some academic belonging, but missing the social and personal dimensions students appear to prioritise. That audit gives teams a faster way to spot blind spots before another survey cycle locks them in.
Second, institutions should pair belonging scales with better open-text prompts. Questions such as "What has helped you feel part of the university this term?" and "What has made it harder to feel like you belong?" are more likely to surface mechanisms that teams can actually act on. Analysing those answers by ethnicity, mode of study, commuter status, or first-generation status can show whether some groups are carrying extra belonging work that averages conceal. The gain is not just richer data. It is a clearer route from feedback to targeted action.
Third, teams should treat belonging as an operational design issue rather than only a communications issue. Peer networks, induction, and welcome week activities that strengthen peer belonging, as well as group work, visible representation, everyday staff responses, and how quickly exclusion is addressed, all shape the social and personal conditions of belonging. For teams using Student Voice Analytics, that means tracking belonging-related comments alongside themes such as respect, support, inclusion, and assessment fairness, then using those patterns to prioritise practical changes. The benefit is that inclusion work becomes more specific, more defensible, and easier to follow through.
If you want a more robust way to analyse belonging-related comments at scale, explore Student Voice Analytics. You can also use our NSS open-text analysis methodology and student comment analysis governance checklist to build a more repeatable approach to belonging analysis.
FAQ
Q: How should universities redesign belonging questions after reading this paper?
A: Start by checking whether your current items cover social, personal, and academic belonging rather than only one of them. Keep a small core scale for tracking over time, but add one open-text question that asks students what has most helped or hindered belonging recently. That gives you both a trend measure and a source of explanation.
Q: What are the limits of drawing conclusions from 69 qualitative interviews in one national context?
A: The study is strong for construct definition, but it does not estimate prevalence in the way a large survey would. The Flemish Belgian setting also has its own institutional and demographic features. For UK practice, the right move is to use the findings as a guide for what to test locally, then check whether the same domains appear in your own surveys, comments, interviews, and focus groups.
Q: What does this change about how we use student voice data on inclusion and belonging?
A: It shifts the task from simply monitoring whether belonging is "high" or "low" to understanding what belonging consists of for different groups. That makes student voice data more diagnostic. Instead of treating comments as anecdotal, institutions can use them to identify which parts of belonging are under strain, and where action is most likely to improve the lived experience.
References
[Paper Source]: Jente De Coninck, Peter Stevens and Wendelien Vantieghem "Sense of belonging defined: how ethnic-minority students conceptualise belonging in the university" DOI: 10.1080/03075079.2025.2507783
Request a walkthrough
Book a free Student Voice Analytics demo
See all-comment coverage, sector benchmarks, and reporting designed for OfS quality and NSS requirements.
- All-comment coverage with HE-tuned taxonomy and sentiment.
- Versioned outputs with TEF-ready reporting.
- Benchmarks and BI-ready exports for boards and Senate.
UK-hosted · No public LLM APIs · Same-day turnaround