Students who frequently consider leaving university report significantly lower sense of belonging than those who don’t, and the effect holds independent of academic performance. That finding, from a predictive retention model built by researchers at St. Cloud State University and published via the National Science Foundation’s public access repository, points to something institutions have suspected for years but rarely measure with any rigor: belonging and academic engagement are separate levers on retention, not the same one wearing two names.
A separate quantitative study on socio-economic status and belonging, published in Teaching in Higher Education and archived via Goldsmiths’ research repository, found that academic engagement and social engagement independently predict belonging and retention — meaning a student can be doing well in coursework and still be at meaningful risk of leaving, for reasons a grades dashboard will never surface.
Quick takeaways
- Belonging and academic performance are independent predictors of retention — strong grades don’t rule out belonging risk.
- First-generation students consistently report lower belonging scores across multiple independent studies.
- A belonging survey needs to separate academic and social dimensions to be diagnostically useful.
Why belonging is now an institutional metric
Belonging used to live in the “soft” category of student experience — important in principle, rarely tracked with the same rigor as enrollment or GPA. That’s shifted as retention data has made the cost of ignoring it explicit. Multiple studies, including research reviewed via the Princeton-affiliated Review of a Sense of Belonging at University, consistently find that students with lower belonging scores are more likely to report considering withdrawal — and that first-generation students, in particular, report meaningfully lower belonging than continuing-generation peers, even when academic performance is comparable.
Designing a survey that captures more than satisfaction
A generic satisfaction survey asks “are you happy here.” A belonging survey needs to ask something more specific and more structurally useful: does this student feel academically integrated, and separately, do they feel socially connected. The research is consistent that these are different constructs with different interventions attached to each. A well-designed instrument should include:
- Academic integration items — comfort participating in class, relationships with faculty, confidence navigating academic support resources.
- Social integration items — peer relationships, sense of fitting into campus culture, participation in student life.
- A direct belonging measure — a validated scale item asking about felt belonging specifically, rather than inferring it solely from the above.
Longitudinal tracking across cohorts
Belonging is not static across a student’s first year — research on transition points consistently shows it dips at predictable moments, often around the first set of graded assessments and again mid-way through the first semester. A survey run once, at orientation, misses both. Institutions getting the most diagnostic value run a shorter pulse version of the same instrument at two or three points across the first year, using QuestionPro’s longitudinal research tools to track the same cohort over time rather than treating each survey as a standalone snapshot.
Turning results into action, not a shelved report
The research is clear that belonging predicts retention risk before a student’s grades typically show it. That only has value if the data reaches an advisor or student success team while there’s still time to intervene — not as an annual report reviewed months after the students in question have already left.
Frequently asked questions
Is student belonging the same thing as student satisfaction?
No. Research consistently treats them as related but distinct constructs — a student can be satisfied with the academic experience while still feeling socially disconnected, or vice versa, and each requires a different intervention.
When during the academic year should a belonging survey be run?
Research points to early-semester and mid-semester as the two most diagnostically useful points, since belonging dips are strongly associated with the first set of graded work and the middle of the first term.
QUESTIONPRO ACADEMIC
Build a belonging survey that separates academic and social integration.
