How belonging changed during first year in two English universities

Updated Sep 07, 2026

Updated, 7 September 2026: Two briefings on this study have been consolidated. We have added the actual study design and limitations, and removed claims that short surveys or comment analysis identify causes or guarantee stronger belonging.

What the study found

David Gilani, Liz Thomas and Daniel McArthur’s Dynamic belonging was published online in Studies in Higher Education on 22 February 2026. It follows 101 first-year students at two English universities through surveys in October, December, February and May; 65 also contributed diaries or open-text responses.

Average belonging declined, with the May comparison against October statistically significant. Early scores were similar across parental-education groups, while differences emerged later. The detailed comparisons distinguished students with no, one, or two university-educated parents; the pattern should not be reduced to a universal first-generation deficit.

Diary analysis suggested possible explanations involving peer connections, cultural barriers, relationships with staff and discussion of finances. It does not establish that those experiences caused the measured change.

Only 1.8% of eligible students participated, and not everyone completed each follow-up. The findings cannot be assumed to describe either institution’s whole population or every first-generation student.

Questions for a local belonging review

Our practical interpretation is to check when your evidence was collected before comparing results. An induction survey and a later module survey may capture different circumstances and different respondents. Keep question wording, timing and participation visible in any report.

If you add an open question, decide what you can act on and how students will hear the outcome. For example, ask what currently helps or hinders connection with their course, then discuss the responses with an appropriate student group. Treat the accounts as experiences to investigate rather than a ranked list of proven causes.

This paper does not validate a one-question belonging scale, an ideal survey frequency or a particular commercial analysis service. Choose a measure suited to your purpose and test whether the resulting evidence supports the decision you want to make. The public NSS methodology illustrates why response coverage and comparability need explicit checks; it is not a belonging measure validated by this study.

Reference

Gilani, D., Thomas, L. and McArthur, D. (2026). Dynamic belonging – how student belonging changes over time for first-generation students. Studies in Higher Education. Open published paper. DOI: 10.1080/03075079.2026.2631772.