Belonging in schools: what it is and how to measure it
This article gives school leaders a working definition of belonging in schools, the research connecting it to pupil outcomes, and practical guidance on measuring it through pupil voice. It includes a worked example of reading results across a trust and a first step you can take this term.
Belonging in schools is a pupil's sense of being accepted, valued and included by the people around them, and it can be measured and acted on like any other performance indicator. It is the pupil's own perception, which is why attendance, behaviour and attainment data cannot stand in for it. This article sets out a working definition, the evidence connecting belonging to outcomes, how to measure it, a worked example of reading results across a trust, and a first step you can take this term.
A working definition, and the ideas it gets confused with
The most cited definition in the research comes from Carol Goodenow's 1993 work in the US, which described belonging as the extent to which pupils feel personally accepted, respected, included and supported by others in the school social environment. The 2018 meta-analysis by Allen and colleagues, published in Educational Psychology Review, drew together a large body of studies and kept that core. For a school leader the practical version is simple: does this pupil experience school as a place where they are accepted, respected, included and supported by peers and adults?
Several adjacent ideas get mistaken for it.
- Happiness is a mood. A pupil can enjoy break time and still feel they do not fit in lessons.
- Good behaviour is compliance. Pupils who follow every rule can feel invisible.
- Attendance is a lagging outcome. It tells you a pupil came in, not how they experienced the day.
- A strong culture, as staff describe it, is an adult judgement. Staff and pupils often see the same corridor differently.
That last gap is the reason belonging needs its own measure. Adults tend to rate school climate more favourably than the pupils living in it, so adult judgement alone misses the experience you are trying to understand.
Belonging is also not a pastoral extra, and it is not only about popular pupils. It is shaped by ordinary things: how a teacher responds to a wrong answer, what happens in the lunch queue, whether a pupil sees their language, culture or family reflected in the curriculum and displays.
It also varies inside one school. Year 7 entrants, pupils with SEND, pupils with EAL, disadvantaged pupils and children who join mid-year can report very different experiences from the same building. A single whole-school figure will not show you that.
Why belonging behaves like a performance and risk indicator
The international evidence points the same way as the school-level evidence. The OECD's PISA 2018 results reported on pupils' sense of belonging at school across participating countries, and found that pupils who felt they belonged tended to show greater engagement and stronger outcomes than those who did not. Allen et al. (2018) found associations between school belonging and academic motivation, engagement and attainment, as well as links to behaviour and attendance. These are associations, not proof that raising belonging alone will move a single outcome. They are strong enough, though, to treat belonging as something to track alongside the usual measures.
Leading and lagging measures
Attendance, suspensions and exclusions are lagging indicators. By the time they move, the pupil's disengagement is usually well established. Perception data is a leading indicator: how pupils feel about school tends to shift before their behaviour does. Satchel Pulse's own analysis of belonging data suggests it can flag cohorts and academies at risk of rising absence, suspensions and exclusions up to 18 months before those patterns show in attendance or behaviour records. That is the company's finding rather than an independently published result, and the lead time will vary between schools. The sensible reading is that belonging gives you an earlier signal, and you should test it against your own data over several terms.
Staff belonging
The same logic applies to staff. Workload and a sense of being valued affect whether teachers stay, and staff turnover affects consistency in the classroom, which is itself part of pupils' experience. The DfE publishes school workforce statistics each year, and they are the right place to check current retention figures for your phase and region before you quote any number to trustees. Collecting staff views with the same method as pupil views lets you see whether the two move together.
Policy context
The government's Every Child Achieving and Thriving white paper, published in February 2026, places weight on inclusion and attendance. Check its current wording before citing it in board papers. For leaders, its relevance is practical: belonging data speaks to both priorities directly.
How schools measure belonging: surveys, questions and timing
Existing data tells you what happened. Attendance registers, behaviour logs and attainment results record events and outcomes, not how pupils experienced them. Headteacher summaries add context but are not comparable across schools, because each head frames their own picture. To compare experience, you need to ask pupils directly, using the same questions in each school.
What a good survey looks like
A belonging survey should take pupils roughly 5 to 10 minutes. Use a validated or research-informed question set rather than writing questions from scratch, and cover the main areas that shape experience:
- relationships with adults
- connection with peers
- feeling safe and treated fairly in daily school life
- having a voice and taking part
- seeing themselves represented in school life
There is no single universal question set, so be wary of any claim that there is. What matters is that you choose one, explain why, and keep it stable so results can be compared over time.
Practical decisions
Anonymous responses tend to encourage candour, while named responses let you follow up with individual pupils and link results to cohorts from your MIS. Many schools use anonymous surveys with cohort tags for exactly this reason. Wording needs to suit the age group, so primary and secondary versions will differ. For EAL pupils, translation and audio read-aloud remove a reading barrier that would otherwise distort the results. Ask pupils, staff and parents consistent questions where you can, so you can see where their views match and where they split.
Timing and response rates
Avoid exam and assessment windows, when pupils are tired and mood is not typical. Run the survey each term, or at the very least twice a year, because one reading shows a position and two show movement. Aim for high participation, since low response makes cohort results unreliable. Be careful with small groups. If a cohort has only a handful of respondents, a single answer swings the average and can identify individuals, so set a minimum group size below which results are suppressed.
Reading the results: cohorts, comparisons and the story behind the average
A whole-school average can look reassuring while hiding the pupils with the weakest experience. Break results down by year group, SEND, Pupil Premium or disadvantaged status, EAL, FSM and combinations of these, such as Year 7 pupils with SEND who are also eligible for Pupil Premium. The gaps between groups usually matter more than the headline figure.
A worked example
Suppose a trust of six secondary academies runs the same survey in the autumn term. The figures below are hypothetical and for illustration only.
Across the trust, Year 7 belonging scores 72 out of 100, which looks acceptable. Splitting by cohort shows that in four academies, Year 7 pupils with SEND score between 68 and 71. In the other two, the same group scores 54 and 57. Nothing in the attendance data for those two academies has changed yet. The free-text comments from those pupils keep returning to the same themes: unstructured times such as break and lunch, and the movement between lessons, where there is no adult nearby and no clear place to be.
That is a specific, actionable finding. It points to two academies, one cohort and two parts of the school day, rather than a vague instruction to improve belonging.
Compare like with like
The comparison only means something because every academy used identical questions at the same point in the term. Differences then reflect experience, not survey design. Treat them as prompts for inquiry and support. A low score is a reason for a conversation with that academy's leaders about what is happening and what help they need, not a league table position.
Use free text to explain the numbers
Scores tell you where to look and comments tell you why. Group comments by theme and sentiment, then check what you find against attendance and behaviour data for the same cohorts. If lunchtime comments cluster among Year 7 pupils with SEND, look at whether lunchtime incidents or part-day absences are also higher for that group. Agreement between sources strengthens the case for action. Disagreement is worth knowing too.
Turning belonging data into action that leaders can track
The value comes from repeating a short cycle each term. Measure, break the results down by cohort, talk with pupils and staff to understand what sits behind them, respond with a targeted action, then re-measure to see whether it worked. A one-off survey gives you a snapshot. The cycle gives you evidence about what changes experience in your setting.
Give each action a named owner and a specific cohort. In the worked example, the SENCO might lead a change to how unstructured time is supervised and structured for Year 7 pupils with SEND, such as a staffed lunchtime club and adults positioned at known transition points. Another academy might start a transition buddy scheme. Whichever you choose, fix the question you will review next term so success is defined in advance.
Tell people what happened. A visible “you said, we did” message to pupils, staff and families shows that responses lead somewhere, and that tends to support participation in the next round. Before you scale an approach across a trust, evaluate it first. The Education Endowment Foundation's guidance on evaluating approaches is a useful reference, so check its current wording and apply it to a small pilot before wider rollout.
For trustees, keep reporting simple and consistent from term to term:
- the headline trend across the trust
- cohorts and academies of concern
- actions in place, with owners
- change since last term
A tool such as Perceptions can handle distribution through your MIS, cohort breakdowns and free-text analysis, so leaders spend their time on response rather than spreadsheets.
One cohort, three or four questions, a named owner
Start small this term. Choose a single cohort or transition point, such as Year 7 in the autumn term. Agree three or four belonging questions drawn from the areas above, set a survey window clear of assessments, and decide who will own the response before the results arrive. That last step matters most, because a result with no owner tends to become a slide in a meeting rather than a change in school life.
Run it again next term with the same questions. Two readings will tell you more than any single score.
Where is your school on the belonging journey?
Take the free belonging self-check. A few quick questions place your school at one of four stages, from exploring to sustaining, and show what to focus on next.
Take the self-check
Related articles
Contact our education team
