Find a twin

Choose a school and find the schools in England most like it in intake and setting, then the look-alikes that do better on absence, attainment or value added.

NoteHow to use it
  1. Choose a school. Pick a region, local authority or district, then a school within it.
  2. Choose what to match on. Matching uses intake and setting only, never results or attendance. By default it weights what matters most for the measure you choose.
  3. Closest twins are the schools most like yours anywhere in England.
  4. Anti-twins are the look-alikes, from the 50 (or 25, or 100) most similar, that do furthest better than yours on the measure you choose:
    • lower absence of a kind;
    • higher value added or Attainment 8;
    • or absence and value added together.

Click a twin, in the chart, on the map or in the table, to compare its fact card with your school’s, including value added and absence year by year. Print for governors gives a clean copy of the comparison for a meeting. The address bar keeps your choice, so a link takes someone else to the same comparison.

How the matching works

Every school is described by the characteristics the absence models use, in groups you can switch on or off:

  • Pupils: disadvantage (FSM), KS2 prior attainment, EAL, White British, joining mid-phase, and the share of girls.
  • Needs: SEN support, EHC plans, and social, emotional and mental health need.
  • Neighbourhood and place: neighbourhood education, and rural or urban.
  • The school: size, admissions, sixth form, and religious character.

For each measured characteristic, the difference between two schools is taken in standard deviations, capped at three. For each category, it is 0 if they match and 1 if not. The average across the characteristics is the distance, and similarity is one minus it, as a percentage. This is a form of Gower’s distance, as in the original project’s twin finder.

Weighting. By default each characteristic is weighted by how much it matters for the measure you chose. A school that matches yours on what drives that measure counts as more alike than one that matches on what does not. The weight is how much the characteristic moves the measure across schools: the standard deviation of its part in a regression of the measure on all the characteristics (log absence for absence). The weights are scaled to average one. Value added and Attainment 8 both take the Attainment 8 weights. Value added is already what is left once intake is allowed for, so the twins should match on what drives results. The combined measure takes the average of the overall absence and Attainment 8 weights. You can switch to equal weights.

Table 1: Matching weights by measure (average 1), and how much of each measure the characteristics explain between schools (R²).
Overall absence Illness Unauthorised Persistent Severe Results (Attainment 8, value added)
Free school meals 2.62 4.16 4.26 2.81 1.96 1.95
KS2 prior attainment 2.46 1.71 1.79 2.54 2.59 3.83
EAL 0.50 0.64 0.16 0.27 0.05 0.29
White British 3.00 2.02 1.57 2.18 3.76 1.41
Joined mid-phase 1.65 0.46 1.31 1.35 1.34 1.20
Girls 0.54 0.65 0.26 0.43 0.72 0.36
SEN support 0.19 0.41 0.05 0.12 0.05 0.06
EHC plans 0.11 0.37 0.26 0.19 0.05 0.15
Social, emotional and mental health need 1.35 0.29 1.11 1.01 1.15 0.62
Neighbourhood education 0.11 0.39 0.48 0.21 0.33 1.45
Size 0.09 0.72 0.49 0.34 0.46 0.63
Rural or urban 0.19 1.07 0.48 0.27 0.20 0.23
Admissions 0.85 1.30 1.87 1.94 1.78 1.85
Sixth form 0.77 0.51 0.36 0.77 0.19 0.68
Religious character 0.57 0.30 0.55 0.56 0.48 0.29
R² of the model 59% 26% 66% 64% 61% 82%

Similar intake is not the same as identical circumstances. A twin can differ in what the data does not see: its pupils’ health, its community, its buildings, its staffing. An anti-twin that does much better is a reason to ask what it does differently, not proof that copying it would work.

Data. The data for each kind of measure is:

  • Absence: autumn and spring 2025/26.
  • Characteristics: the January 2026 school census and SEN data.
  • Results: Attainment 8 for 2024/25 and Value Added v2.1 over 2021/22 to 2024/25, shrunk, each for all, disadvantaged and non-disadvantaged pupils.
  • Absence against similar schools: within the school’s own authority, as on the absence page.
  • Over time: each year’s value added against England (unshrunk; the latest year provisional), and absence in the autumn and spring terms from 2016/17, followed through academy conversion and changes of trust as on How school absence changes.

All of it is published school-level data.