All resources · 2,296 words · 10 minute read

Parent and student surveys as school inspection evidence

How a school parent survey becomes inspection evidence, not a percentage: question design, defensible response rates, and the survey your authority ran.

A survey becomes evidence when the same question is asked twice with a decision in between. Until then it is a percentage: probably accurate, occasionally interesting, and of almost no use to somebody whose job is to check rather than to believe. Everything below is about arranging a year so the second asking exists and the decision is on the record.

Most schools here run at least one parent survey, and many now also sit inside a survey their authority runs and they do not control. Both are useful. Both are wasted the same way, by being reported and then filed.

What makes a school parent survey inspection evidence rather than a number

Our piece on what counts as evidence in a school inspection sets out the general properties: dated, attributable, showing change, findable by a stranger. A survey is dated automatically, shows change if you run it twice, and is findable if you store it sensibly. What it cannot do is name anybody, because you promised it would not, and anonymity is the condition on which people answer honestly.

So the attributable unit of a survey is not a person. It is the cohort and the instrument. A reviewer can check that you asked Grade 9 parents, in November, this exact question, and how many answered. That is attribution enough, provided you kept the wording, the dates and the denominator.

Which means the record you keep is not the chart. It is four things: the questions as asked, the population you asked, the date, and the raw distribution of answers. Everything else can be rebuilt from those, and nothing can be rebuilt without them.

The middle term, which is the part schools lose

One survey is a photograph of a mood. Two are a photograph of two moods. What makes the pair an argument is the thing between them, and that is where the record usually stops.

In November a parent survey says communication about progress is the weakest area. In May the same question comes back three points higher. Presented like that, a reader cannot separate improvement from weather: parents in May have just had a report and are in a better temper than parents in November.

Now add the middle. On 4 December the leadership team discussed the November finding and decided two things: subject teachers would reply to parent emails within two working days, and the mid-year report would carry a comment on next steps rather than a grade alone. The deputy head owned both. In March a sample of emails was checked against the two-day rule. Then May.

Same two numbers, completely different document. The middle takes about twenty minutes a term to write, and its absence is the commonest reason survey evidence does not survive a question.

The pattern is the one that governs every other kind of evidence: a starting point, an action, a later look, held together by dates. The survey supplies the two ends; only the school supplies the middle.

The survey your authority ran, which you did not commission

Part of the evidence base about your school is collected from people who do not work for you, and it is collected before anybody arrives.

In Dubai, the Schools Inspection Bureau's Initial Quality Inspections Handbook, Edition 1 carries a short section headed "Evidence from parents". Each school is asked to distribute a Parents' Questionnaire to a sample of families whose children attend, the size of that sample depends on the total number of students on roll, and a summary of the issues raised forms part of the inspection evidence base. Note who distributes it: the school does, to its own parents, and the answers go somewhere else.

KHDA has separately published a Dubai Student Wellbeing Census. Its FAQ page, last updated on 22 November 2021, invites all students in Grades 6 to 12 at Dubai private schools to take part, runs from 31 October to 30 November, and reports to schools in February. Answers are grouped so schools cannot see individual responses. Two sentences sit together there and both matter: schools are encouraged to use the results as part of their self-evaluation process, and "Inspectors will have access to each school's report."

That page describes the fifth year of a five-year project begun in 2017 and has not been updated since, so whether the census still runs in that form is not settled by KHDA's published pages. Check what your own authority sends you. The principle outlives the uncertainty: where an authority collects a view of your school and reads it back to itself, your February is not free.

Abu Dhabi runs its own. ADEK's Abu Dhabi Parent Survey page describes an annual survey of parents across private, public, Charter and SoD schools and nurseries, used to develop policies, programmes and services, and says the department also gathers insights from students and school staff. Qatar News Agency reported in March 2023 that Qatar's education ministry had begun a parental opinion poll for 2022-2023 covering parents from kindergarten to secondary; the ministry site could not be reached to confirm the detail, so treat that as reported rather than as specification.

Find out what your authority asks your community and on what schedule, then put the date it reports back to you in the calendar as a working meeting rather than a notification.

Question design, or why "are you satisfied" is a wasted line

Satisfaction questions produce numbers that move for reasons you cannot trace. Behaviour questions produce numbers attached to something you can change on a Tuesday.

The test is whether an answer names a decision. "How satisfied are you with communication?" names nothing. "The last time you contacted the school with a concern, how many working days passed before somebody replied?" names a target, a person and a process, and a school that shortens the reply time sees it move for the right reason.

Two substitutions worth making, in either language:

• Not "do you feel informed about progress" but "can you name the one thing your child's teacher wants them to improve this term".
• Not "is the school safe" but "if something worried your child at school, who would they tell".

Keep one satisfaction question if the board wants a headline. Then two rules of construction: never change the wording of a question you intend to compare, and always keep one free-text box, because free text is where you find out which group the percentage is hiding.

Response rates, and the bad one

No Gulf authority publishes a threshold that makes a school survey acceptable, so treat the rate as something you explain rather than something you pass.

Do the arithmetic before you send anything. A school of 900 students has roughly 600 families; a 30 per cent return is 180 responses, plenty for a whole-school reading and thin for a claim about Grade 11. Whether a rate is defensible depends on what you intend to say with it, and most schools decide the claim after seeing the number, which is the wrong order.

What raises a rate is unglamorous: a short instrument, a stated purpose, a deadline, one reminder, and proof from last time that answering changed something. The last of those compounds, and it is the only one still working in year three.

With a low rate, report it and analyse the shortfall: which year groups are missing, and whether the Arabic version went out at the same hour as the English one or three days later. A 28 per cent response rate with a paragraph explaining who did not answer is worth more than 61 per cent presented without comment.

When the answer is unwelcome

Every school eventually gets a result it did not want, usually in an area leadership believed was a strength. Three responses are common and two are mistakes.

The first is to explain it away in the report: small sample, unhappy parents, ambiguous question. Some of that may be true. All of it reads as a school arguing with its own data, and a reader who sees that stops asking about communication and starts asking about leadership.

The second is quieter and worse. Nothing is announced, the instrument simply does not go out the following year, and the record shows a school that measured something once. That pattern is legible from outside, and it is a governance question rather than a survey question.

The third works, and it is arithmetic rather than courage. A poor November reading, a dated decision and a better May reading is the strongest evidence available in this subject. A good November reading is a pleasant slide with nowhere to go. Where you are judged on the honesty of your self-evaluation, and in several systems here you explicitly are, the bad number is the asset.

Write it in as a weakness with an action attached rather than a strength with a caveat, which is the harder half of writing a school self-evaluation form reviewers trust.

A twelve-month cycle that produces evidence

1. Term one, week three. Run the parent survey early, while the year is still changeable. A survey in May arrives too late to act on.
2. Within two weeks. Read the free text first and the percentages second. Pick one finding. One.
3. Within a month. Take a dated decision on it with a named owner, and minute it. This is the middle term, and it is the whole exercise.
4. Mid-year. Check whether the decision happened in classrooms, not whether it was announced. Record what you found, including when the answer is no.
5. February, or whenever your authority reports. Open the authority-run report the week it arrives and compare it with your own. Where the two disagree, that is the interesting finding.
6. Term three. Ask the identical question. Do not improve the wording.
7. Before the year ends. Publish a short note saying what you asked, what you heard and what changed. It costs an afternoon and buys next year's response rate.

Seven steps, perhaps six hours across a year, against the effort most schools spend producing a report nobody acts on.

Three ways survey evidence falls apart

The instrument changed. Somebody improved the questions between waves and the comparison is gone. Freeze the core questions for three years and add new ones at the end.

Only the summary survived. The chart is in the folder; the wording, the distribution and the denominator are on the laptop of somebody who has left.

Nobody can say what changed. The survey said communication, the report said communication, and no teacher can name one thing the school did about it. That finding is not about surveys, and it surfaces in the interviews described in what happens during an inspection visit week.

For the year-round rhythm that makes surveys an ordinary input rather than an annual event, start from our guide to school inspection readiness in the Gulf.

Frequently asked questions

What makes a parent survey count as inspection evidence? Asking the same question twice with a dated decision in between. One reading is an opinion poll. Two readings separated by months, with a record of what the school changed and who owned it, is an argument a reviewer can check. Without the middle term, a percentage proves only that you asked.

What is a defensible response rate for a school survey? No Gulf authority publishes a threshold, so treat the number as something you explain rather than something you pass. Report the rate, say which groups are under-represented, and say what you did to reach them. A low rate you have analysed is more credible than a high one you have not.

How should we use a survey our education authority ran? Open it the week it arrives, choose one finding, act on it with a named owner and a date, and keep the record. KHDA's published census FAQ says schools are encouraged to use results in self-evaluation and that inspectors will have access to each school's report, so the gap between receiving it and using it is visible.

What do we do if the survey result is bad? Keep it. An unwelcome result is the only kind that gives you a credible starting point, and a school that shows a poor reading followed by an action and a better one has made the strongest case available to it. A result quietly withdrawn is a governance finding waiting to be made.

How Asabya Academy helps

Stakeholder surveys in Asabya Academy are built to be repeated rather than run: the instrument, the population and the date are kept with the results, so the same question can be asked again a term later and the two readings sit beside the decision taken between them, with its owner and review date, against the standard it speaks to, in Arabic and English.