Schools collect a great deal of data and act on very little of it. Much of what is gathered exists to demonstrate that something was measured. A smaller set of numbers tells a school whether it is getting better, and separating the two is most of the work.
Start with the question, then find the data
Schools commonly begin with whatever data they already hold, since attendance, grades and survey results are to hand, so they get examined, and a conclusion is drawn from whichever of them moved.
Starting from the question works better, and the question is what the school is trying to find out, at whole-school level, at class level and for an individual child. Is the concern academic progress, or whether students are reaching their potential, or how they are developing outside the syllabus? Rushing to find data that later turns out to be misleading is a common and expensive error.
A number means nothing until you know what sits behind it
GSE worked with a school whose literacy results had improved year after year. Then one year they fell sharply, and the leadership team was ready to abandon a strategy that had been working and make wholesale changes.
Nobody had looked at the intake that year, and the school had taken in a considerably higher number of transient students that year, so the cohort being measured was no longer the cohort that had produced the earlier results, which left a sound strategy looking like a failed one.
Turnover is high in international schools, and it distorts any year-on-year figure in the same way. Before a school concludes anything from a change, somebody has to establish whether it is comparing the same children.
Measure what the school changed
Raw attainment tells you a good deal about a school’s intake and much less about its teaching. Higher socioeconomic groups perform better academically wherever they are enrolled, and a selective school reporting excellent results may be reporting its admissions policy.
The honest measure of a school is the movement it produces. What did this child know in September, what do they know now, and how much of that difference did the school make? A teacher who moves a student from a C to a B has done something demonstrable. A school where able students succeed in spite of the teaching has done considerably less.
This shows most clearly with the average student, because strong students often succeed anywhere, and students in difficulty attract attention and support, so what happens to the large group in the middle is the fairest test of a school.
Opinion is data
GSE holds a position here that many will disagree with, which is that the evidence a school needs does not all convert into tables. Climate, tone and culture are measurable in the sense that matters, since a school can tell whether they are getting better or worse, and they indicate quality as reliably as any percentage.
Two cautions apply to that position, the first being the loud minority, meaning the individual or small group who present themselves as speaking for everybody and are usually speaking for themselves, since a school responding to volume will keep making decisions for a handful of families.
The other is that the people who volunteer their views to a school are a self-selecting group. GSE uses street surveys for this, asking people in the surrounding community what they know or believe about a school. The answers from people with no reason to flatter are frequently the most useful data a school gets, and they are the only way to hear what the families who did not enrol are saying.
Hold conclusions in draft
Every conclusion drawn from data should be treated as provisional until the alternatives have been considered. In practice a leadership team forms a view early, which is unavoidable, and the discipline lies in keeping that view in draft, ready to discard as more comes to light.
Confirmation bias is the danger, and in a school it looks like gathering further data to support the position the head already holds. The first impression is often the opposite of the final determination, and a team that cannot say what would change its mind has stopped analysing.
Separate the data that develops people from the data that judges them
What a set of information is for matters more than any other distinction here, because the same observation used for two purposes produces two different results.
GSE once ran a peer coaching programme in which teachers volunteered to observe each other, then shared feedback and ideas. One rule made it work, which was that nothing from those observations could be passed to administration or outside the trusted group.
Our goal was to improve the quality of teaching overall, not to gather data to judge performance.
Greg Parry, Co-Founder and CEO, GSE
Without that rule the observations would have become performances, with teachers showing their best lesson, avoided anything difficult and said little of use to each other, and the school would have collected a set of numbers describing a version of teaching that does not exist.
Both kinds of information are necessary, and they work differently. Appraisal data judges, and it should be open, fair and known to everyone. Developmental data helps a teacher improve, and it only works where they are certain it will not be used against them. Blurring the two costs a school the second kind, which is the kind that changes teaching.
What to do once you know something
Data that changes no practice is a form of administration. Where a school establishes a real weakness, the response has to reach the classroom, and the people who operationalise it are teachers rather than the leadership team that commissioned the analysis.
That usually means professional learning, and it sometimes means resources, although equipment alone rarely improves an outcome and should not be the first answer. More often it means removing the obstacles that stop teachers acting on what the school has learned, and making the priority visible enough that it appears in meetings, in planning and in what the school talks about.
Everything a school measures should end in a decision about teaching or about how the school runs. Where a set of figures cannot be traced to a decision somebody made differently, it is worth asking why the school is still collecting it.
Read next
- The School Dashboard: What an Owner Should See Every Month. Which indicators move first, and which arrive too late to act on.
- School Meeting Minutes. Evidence a school already produces and never reads.
- Give Teachers Back Their Schools. Why developmental information has to stay with teachers.
- Vision and Mission Integrity. Testing whether stated purpose shows up in decisions.
Part of GSE’s What We Believe collection.
To talk about your school project, reach us through the contact page.