How to build an attendance improvement plan using belonging data

An attendance improvement plan works best when it addresses why pupils stay away, not only how often. This guide shows you how to build a termly plan that places pupil, family and staff perception data beside your attendance figures, so each action is aimed at a cause in a named cohort. Before you start, have three things ready: your latest attendance data by cohort from your MIS, a named senior lead who will own the plan and enough time to run one short survey this half term.

Step 1: Set out your attendance baseline by cohort

A whole-school attendance figure can hide where the problem actually sits. Breaking overall and persistent absence down by year group, and by SEND, Pupil Premium or disadvantage, EAL, FSM and children in care, gives a much clearer picture. Suspensions by the same groupings add useful context where they are relevant. Where numbers allow, combinations are often the most revealing, for example Year 8 pupils with SEND who are also eligible for Pupil Premium.

Definitions matter before any comparison is made. The persistent absence threshold and the rules on authorised and unauthorised absence are set nationally and have changed over time. The DfE’s attendance guidance and the latest “Pupil attendance in schools” statistics release are the reliable reference points. Consistent MIS coding across every academy in the Trust matters just as much, because differences in coding practice can easily look like differences in attendance.

Pupil counts belong alongside every percentage. In a cohort of 12, one child becoming persistently absent shifts the figure by more than eight percentage points. Percentages on their own can draw attention to small groups that moved by chance, while larger groups with steady, moderate gaps go unnoticed.

The result is a one-page table with a row for each cohort and columns for pupil count, overall absence, persistent absence and, where relevant, suspensions. Each row can then be compared with the school or Trust average and with the same cohort the previous year. Three or four priority cohorts usually emerge for further exploration. They are best chosen on a mix of gap size and the number of pupils affected, rather than gap size alone.

Step 2: Add the perception data that attendance figures cannot show

Attendance is a lagging measure. It records absence after a pupil has already stepped back. Belonging, feeling safe in school, relationships with staff and engagement with lessons can shift much earlier, and they are what a pupil weighs up on a Sunday evening. Perception data brings those conditions into view while there is still time to act on them.

Decide who to ask

Pupils are the natural starting point, because they can say directly how school feels. Families add a different angle, as they see the reasons for absence at home and know how contact from school is received. Staff complete the picture. Their views on workload, behaviour climate and capacity to follow up absence show whether a plan is realistic to deliver.

Choose a repeatable question set

A short set of questions covering connection to peers, relationships with adults, safety and engagement with learning works well. Keeping the wording and timing the same each term is what makes the results comparable, since a change in score only means something if the question and the moment were the same. Perceptions by Satchel Pulse offers a validated 5 to 10 minute question set and breaks results down by the same MIS cohorts used in Step 1, so there is no need to rebuild those groups by hand.

Protect your response rate

Surveys completed in a timetabled lesson or form time get far better response rates than those set as homework. It also helps when pupils and parents know beforehand what the survey is for, who will see the results and what will happen next. Response rates fall quickly when people suspect nothing will change.

Step 3: Compare belonging and attendance side by side

Setting each priority cohort’s belonging results next to its attendance figures is where the two sets of data start to explain each other. Take a secondary academy where persistent absence in Year 7 is higher among Pupil Premium pupils. The survey shows the same group scoring lower on belonging and agreeing less often that “there is an adult in school I can talk to”. That pattern points towards relationships and transition rather than punctuality or family engagement, and it leads to very different actions.

Most cohorts fall into one of three patterns:

• Low belonging and low attendance: usually the first priority. These pupils have already stepped back, and the conditions behind it are visible.
• Low belonging, acceptable attendance: worth watching closely. Attendance may follow, so light-touch action now can make a difference.
• Low attendance, good belonging: often a sign of barriers outside school, such as travel, caring responsibilities or health appointments, where working with families is likely to help most.

The link is a signal to investigate rather than proof of cause. A low belonging score shows where to look, not that belonging produced the absence. Any claim about how far ahead belonging moves compared with attendance needs a named source and year, or several terms of the Trust’s own data behind it.

Comparing academies across a Trust works best where the questions and timing were identical. Context matters before drawing conclusions: intake, size, and whether pupils travel long distances in rural areas or short ones in cities. A gap between two academies may reflect context as much as practice. The Trust-wide view is most useful for deciding where central support should go first, and which academies have practice worth sharing.

‍

Step 4: Find the reasons behind the numbers

Scores show where a problem sits, targeted questions and conversations show what it is. Responses from priority cohorts tend to cluster around familiar themes: friendships, particular lessons or subjects, transport, feeling unnoticed, behaviour in corridors, or how school contacts home about absence.

With hundreds of responses, cohort analysis can help. Reading a sample of the raw responses behind each theme confirms that the label fits what pupils responded with and often surfaces ideas that cut across several themes.

Conversations add the detail that responses cannot. A small focus group with pupils from a priority cohort works best when it is led by an adult they already trust, who need not be the person who owns the plan. Three things are worth settling beforehand:

1. Whether responses will be anonymous, and how to make sure no one is identifiable in what is shared.
2. What will happen with what is heard, and when pupils will hear back.
3. Which questions need answering, so the conversation stays on experience of school rather than individual cases.

Conversations with parents follow the same principles. A short call or coffee morning with families in a priority cohort often explains a theme the survey only hints at.

This step ends with a decision about which two or three root causes to address this term. A plan that tries to fix everything dilutes ownership and makes it hard to see what worked. The strongest choices are usually the causes that affect the most pupils in the priority cohorts and that staff have the capacity to change.

‍

Step 5: Write the plan with owners, actions and measures

A single page per priority cohort keeps the plan focused and easy to review. Each row covers the same seven fields:

• Priority cohort
• Root cause
• Action
• Named owner
• Start date
• Review date
• Success measure

Every action links back to a cause identified in Step 4. Where Year 7 pupils say they know no adult well, transition support and a key adult for each pupil in the cohort is the natural response. Where one subject is named repeatedly, a lesson-experience review with the head of department, framed as a shared question about what pupils are experiencing, keeps the focus constructive. Where families say they feel blamed by first-day-of-absence calls, the contact approach itself is the thing to revisit, including the wording, who makes the call and what support is offered.

Each action works best with two measures. One is a perception measure, such as the cohort’s belonging score or its “adult I can talk to” agreement rate the following term. The other is an attendance measure, such as persistent absence for the same group. Belonging can move before attendance does, so a plan judged on attendance alone may look like it is failing after six weeks when the underlying conditions are improving.

Realistic targets describe what a good outcome looks like at two points. After one term, a change in the perception measure and early signs in the attendance data are a reasonable expectation. After a year, the aim is a sustained narrowing of the gap. The Education Endowment Foundation’s attendance guidance summarises what is known about different approaches and is a sound basis for setting modest, defensible expectations rather than promising gains the evidence cannot support.

‍

Step 6: Share results and actions with staff, pupils and families

Reporting back within a few weeks of the survey, in a “you said, we did” format, shows pupils and families that their views led somewhere. Two or three things they raised, each paired with what is being done about it, is usually enough. Not closing the loop is one of the main reasons response rates fall in later surveys, and a lower response rate weakens every comparison the following term.

With staff, cohort-level findings work better than individual comments. Framing priorities as shared, for example “Year 7 pupils with Pupil Premium tell us they lack a trusted adult” rather than “these departments score badly”, makes a real difference to how they land. Staff who feel scrutinised tend to explain the data away, while staff who see a shared problem help to solve it. Including staff survey findings on workload and capacity also keeps the plan realistic about who can do what.

A one-page briefing gives governors or trustees what they need: the baseline, the key perception findings, the actions with owners, and the measures that will show progress. It also gives the board something it can follow term after term and question usefully.

Across a Trust, it helps to agree in advance what central teams see and what stays at school level. A common arrangement is for the central team to see cohort-level results and plans for every academy, while pupil comments and focus group notes stay with the school. Writing this down and sharing it with headteachers before the next survey avoids misunderstandings later.

‍

Step 7: Review each term and adjust the plan

Repeating the survey at the same point each term, with the same questions, allows each cohort to be compared fairly with its previous results. A few checks help before reading too much into any change. In a small cohort, a handful of pupils can explain the whole movement. A shift in response rate can change the picture on its own. The time of year also plays a part, as autumn and summer scores often differ for reasons unrelated to any action taken. Patterns that hold across two or more terms are a much firmer basis for changing course.

Each action then comes up for a decision to continue, change or stop it. Recording the reasoning alongside the decision is what gives the plan its value over time. An action that stopped because it worked reads very differently from one that stopped because it was not delivered. Over several terms, these notes build an evidence trail that governors, trustees and inspectors can follow, showing what was seen, what was done and what changed.

Cohorts that have recovered can come off the priority list and noting this in the plan makes the recovery visible. New cohorts surfaced by the data, including groups that did not appear in the first baseline, can take their place. Refreshing the Step 1 table each term with current figures and pupil counts keeps priorities following the data rather than last term’s assumptions.

Starting with one cohort before half-term

You do not need to build the whole plan at once. This term, pick one priority cohort from your baseline table, schedule a short survey for it before half-term, and book a review meeting with a named lead to turn the results into the first draft of your plan. Once that cycle has run, add the next cohort and repeat. Learn more about our services