Evidence and methods
How Pupil Outlook reaches its numbers
Every figure in Pupil Outlook is one of three kinds: your school's data, published research, or a model estimate. This page explains each, shows the arithmetic, and lists the research with links so you can check it yourself.
Three kinds of figure, always labelled
| Label | What it means | How to use it |
|---|---|---|
| Your data | Figures your school enters: attendance, persistent absence, reading, numeracy and behaviour for the year group and its groups, plus the number of pupils. | Facts. Check them against your MIS. |
| Research | Published findings, mainly the Education Endowment Foundation (EEF) Teaching and Learning Toolkit, the Youth Endowment Fund (YEF) Toolkit and Welsh Government guidance. | Averages across many schools. A guide to what tends to work, not a promise for yours. |
| Model estimate | Pupil Outlook's planning model: how your figures might move with or without each support. | For comparing options side by side. Not a forecast. |
Turning a recommendation into days, pupils and hours
Percentages are hard to picture. Pupil Outlook converts them using simple arithmetic you can check.
per pupil a year for each percentage point of attendance. Welsh schools must open for at least 380 sessions, or 190 days, a year, and 1% of 190 is 1.9.
pupil-days a year for 1 percentage point across a year group of 180 pupils (180 × 1.9). This is a made-up example to show the arithmetic.
a school year (190 ÷ 5). Staff time is shown as hours a year: staff-time units × hours a week per unit × 38. You set the hours per unit.
Ranges, not single numbers
Each model estimate comes with a low-to-high range. The range is wider when the supports in a plan have weaker evidence behind them: roughly 60% to 125% of the likely figure for strong evidence, 35% to 140% for some evidence, and 0% to 150% where no robust evidence was found. These widths are a Pupil Outlook convention to show uncertainty honestly. They are not taken from a research paper.
Old prices in today's money
Where research gives a cost from an earlier year, Pupil Outlook shows it twice: the original price and its value today, adjusted with the Office for National Statistics consumer prices index (CPI). From the 2016 average (about 100.7) to August 2026 (143.6) prices rose about 43%, so a cost of £6 per pupil in 2016 is about £8.60 now. Each price sits next to what it bought, such as less absence or extra months of progress. Where a source doesn't say which year its prices are from, we show the figure as given and say it isn't adjusted. ONS CPI index (D7BT)
Three futures
For the recommended option, the summary shows three paths for attendance: nothing changes, implemented well, and implemented poorly. The EEF's guide to implementation (2024) makes the point that a promising approach's real effect depends on how well it is carried out day to day. "Implemented poorly" assumes 40% of the planned effect. That figure is our illustration, not a research finding.
Leadership frameworks built in
Familiar ways of planning and reviewing, so the summary fits how your senior team already works.
- 1
Implementation stages
Each recorded decision carries one of the four phases in the EEF's guide to implementation: Explore, Prepare, Deliver and Sustain.
- 2
Plan, Do, Study, Act
Every decision has a review date. The decision is the plan, the half-term is the doing, the next figures are the study, and keep, adjust or stop is the act. The cycle comes from quality improvement (Shewhart and Deming).
- 3
Universal, targeted, specialist
Each support shows who it is for: every pupil, groups with an identified need, or a few pupils with specialist help. This is common practice in Welsh schools rather than a formal Welsh Government model.
- 4
Logic model
The development plan draft sets out each priority as inputs, activities, outputs and outcomes, so it is clear how staff time turns into the change you want.
The research we use
"Months" means the average extra progress over a year, as the EEF reports it. Toolkit figures are updated over time, so open the source before quoting a number.
| Approach | What the research found | Strength and cost | Source |
|---|---|---|---|
| Reading comprehension strategies | +6 to +7 months (EEF summaries differ; check the live page) | Moderate evidence, very low cost | EEF Toolkit |
| Small group tuition | +4 months | Moderate evidence, low cost | EEF Toolkit |
| One to one tuition | +5 months | Moderate evidence, moderate cost | EEF Toolkit |
| Metacognition and self-regulation | +8 months | Very low cost | EEF Toolkit |
| Feedback | +6 months | Extensive evidence, very low cost | EEF Toolkit |
| Parental engagement | +4 months; the EEF notes far less research in secondary schools | Very low cost | EEF Toolkit |
| Social and emotional learning | +3 months on average, +5 in secondary | Moderate evidence, very low cost | EEF Toolkit |
| Behaviour interventions | +3 months | Moderate evidence, low cost | EEF Toolkit |
| Mentoring (attainment) | +2 months | Moderate evidence, moderate cost | EEF Toolkit |
| Mentoring (violence) | Violence down about 21% on average across studies | Rated moderate impact; £845 to £3,500 per mentee (price year not given, so not adjusted) | YEF Toolkit |
| Homework | +5 months in secondary | Very limited evidence, very low cost | EEF Toolkit |
| Arts participation | +3 months | Moderate evidence, very low cost | EEF Toolkit |
| Physical activity | +2 months | Extensive evidence, very low cost | EEF Toolkit |
| Attendance approaches | Review of 72 studies (all but three from the USA): communicating with parents and responsive, targeted approaches looked most promising | Evidence rated weak overall | EEF |
| Texting parents (trial, 2016) | Secondary schools in England: absence fell; maths about +1 month. Cost about £6 per pupil per year at 2016 prices, about £8.60 in August 2026 money | Single large trial | EEF project evaluation |
| Reciprocal Reading (trial, 2019) | 98 primary schools: +2 months for struggling readers in Years 5 and 6; no impact for the whole-class version | Primary only | EEF project evaluation |
| Magic Breakfast (trial, revised 2019) | +2 months in Year 2; no evidence of an attainment effect in Year 6 (re-analysed in 2019 as non-randomised) | Primary only | EEF |
| Pupil Development Grant | £1,150 per eligible learner in 2026–27; spending must be based on evidence of impact | Welsh Government funding | Welsh Government |
Figures were checked against the publishers' own summaries in October 2026. Pupil Outlook is independent and is not endorsed by the EEF, the YEF or the Welsh Government.
What Pupil Outlook can't tell you
- It can't predict the future. Model estimates show how options compare, not what will happen.
- Research averages hide wide variation. Your school's context and how well you implement matter more than the average.
- It doesn't know about individual pupils. It works with group totals only.
- It doesn't replace professional judgement. Your senior team decides.
See it with example figures
The demo runs in your browser with a made-up Year 8. Nothing you type is sent anywhere.