Belonging is the factor most tied to wellbeing, and it is won early

The factor closest to wellbeing

In 2026, EVA (the Danish Evaluation Institute) set out four factors that universities can act on to strengthen students’ academic outcome, wellbeing and completion: academic engagement, self-efficacy, collaboration and belonging. The fourth is belonging.

Belonging is a student’s sense of being part of their programme and their peers, of feeling recognised in an academic community. EVA measured the social-connection side of that specifically. Of the four factors, belonging is the one most tied to how students feel.

What it changes

-9 pts lower dropout risk, high vs low belonging

+0.3 higher grade average

-16 pts lower risk of poor wellbeing

EVA 2026 · high vs low on this factor, corrected for background

Those are EVA’s numbers, not ours. They describe what happens to completion, grades and wellbeing when belonging is high rather than low. EVA is explicit that belonging matters most for wellbeing, and that it is formed most in the transition into study life.

Why the timing matters

The second half of that finding is the useful part. Belonging is largely won or lost in the first weeks of a programme. A student who arrives, finds people, and feels recognised builds a base that carries through the hard stretches. A student who arrives and stays on the edge starts from behind, and the gap tends to widen.

That makes belonging different from a factor you only notice once something has gone wrong. It is visible early, and it responds to the frames an institution sets around the study start. The window to act is at the beginning, not after a withdrawal.

The research names the factor. The Framework moves it.

EVA tells you belonging matters. It does not tell you where you stand on it, in which programme, this month. StudentPulse measures belonging continuously, through the topic that captures it:

In StudentPulse today

7.9 Sense of Belonging & Community

Sector means, 0 to 10

And the pattern in our own data points the same way as EVA’s wellbeing finding. Students with a strong sense of belonging report wellbeing close to two points higher than those who feel they do not belong. We set out that full correlation, and how belonging buffers academic stress, in a companion piece: belonging buffers academic stress. The point here is narrower. EVA places belonging inside a frame of completion and outcome, and names it the factor closest to wellbeing. The research tells you which factor matters. The Framework is how you watch it move.

What works, where belonging is high

Belonging is one of the highest-scoring things students report, with close to seven in ten responses in the top band. The high comments are strikingly consistent, and rarely about events or societies. They describe a class where people help each other and talk well, where teachers deliberately build the connections, and where there is still room to be yourself.

What it means for institutions

1. Design the study start as belonging work. EVA finds belonging is formed most in the transition into study life. An inclusive study start, with structured chances to meet and be recognised, does more than a welcome week. Measure belonging from the first weeks so you can see whether the design is working, not assume it.

2. Build the frames for relationships. Belonging is not left to chance. Timetabling, how classes are formed, and structured group work all set the conditions for students to connect. Support student communities and voluntary activity, and give students real influence over their physical study environment.

3. Watch the students most at risk of not belonging. First-in-family students, students with a disability, and minority students are on average more likely to feel outside the community. A short check-in surfaces who is drifting while there is still time to route a nudge toward a study group, a peer mentor, or a conversation, and to close the loop so students see that answering changed something.

How we measure it

Belonging is captured through the Sense of Belonging & Community topic, scored 0 to 10 across short check-ins. The companion correlation uses a paired, within-respondent design: students who answered both a belonging item and a stress or wellbeing item in the same submission. Those relationships are associations, not proof of causation. Cross-institution aggregate; no single institution identified; minimum group size applied. Student comments are de-identified, translated where needed, and drawn from the highest-scoring responses on the topic. The four factors and their effect sizes are from EVA (the Danish Evaluation Institute), 2026.