Key finding: the education data a UK dissertation can realistically use is public and aggregated, not pupil-level. The Department for Education’s Explore education statistics service, Ofsted’s inspection data and HESA’s open higher education data can be downloaded and analysed within a term. Pupil-level records in the National Pupil Database are restricted to approved researchers and are not a realistic source for an undergraduate or taught master’s project.
Which education data sources can you use, and how do you get them?
| Source | What it covers | Access, from the publisher’s own pages | Good for | Not good for |
|---|---|---|---|---|
| Explore education statistics (Department for Education) | School and college performance, pupil statistics, educational workforce, further and higher education, school absence and attainment, in England | Statistical summaries, a data catalogue of open data files, a data table builder and an API; content is available under the Open Government Licence v3.0 | School-level and area-level comparisons, trends over time, workforce questions | Individual pupils; anything outside England |
| Ofsted state-funded school inspections: management information | Inspection outcomes for state-funded schools | Monthly releases as ODS and CSV files, including latest inspection outcomes and a year-to-date file; the release we checked was published on 10 September 2026 with data as at 31 August 2026 | Comparing inspection outcomes with school characteristics, region or time | Explaining why a school received a judgement |
| HESA open data on higher education students | Detailed extracts from the student data collection for UK higher education providers | Published as open data under a Creative Commons licence (CC BY 4.0), with interactive tables that can be downloaded in machine-readable form | Questions about student numbers, subjects, personal characteristics and providers | Individual student experiences or outcomes beyond what the tables include |
| National Pupil Database and other DfE personal data | Pupil-level records | Shared by default through the ONS Secure Research Service, which requires ONS researcher accreditation, with some data via the UK Data Service; direct access through the DfE Data Sharing Service is considered only if you meet one of three exemptions (an essential public task, research funded or commissioned by DfE or another government department, or research sponsored or supported by DfE); outputs must pass statistical disclosure control | Funded or sponsored projects with a secure-access route | A student dissertation, unless your supervisor can show an eligible route |
Education is devolved, so the sources above mostly describe England. If your dissertation is about Scotland, Wales or Northern Ireland, look for each administration’s own published statistics and say clearly which nation your data covers. Our guide to UK data sources by subject lists wider options.
Which source fits which research question?
| If your question is about… | Start with | Typical analysis |
|---|---|---|
| Absence, attainment or workforce patterns across schools or areas | Explore education statistics | Comparisons between groups or regions, trends over time, correlation |
| Inspection outcomes and school characteristics | Ofsted management information joined to DfE school data | Cross-tabulation, chi-square, logistic regression if you have the training |
| Higher education participation or subject choice | HESA open data | Descriptive trends, comparisons between subjects or providers |
| Experiences, attitudes or practices | Your own data collection | Surveys or interviews, with national data as context |
The key discipline is matching the unit of analysis to the question. Schools and local areas are the usual units in public data, so write questions about schools and areas, not about individual pupils. Our guide to the education dissertation methodology chapter explains how to justify a secondary-data design.
How do you download and prepare the data? A short procedure
- Write the question first, including the unit of analysis and the time period.
- Find the dataset in the data catalogue and download the data file together with its guidance or methodology note.
- Record the source details: publisher, title, release date, the period covered and the date you downloaded it. You will need them for your reference list.
- Check definitions. Variables can change names, categories or coverage between years, so read the notes before you combine files.
- Check suppression. Published figures for small numbers are often suppressed or rounded, so decide how you will treat missing cells and report it.
- Clean in a copy. Keep the original file untouched, and write down every cleaning step so someone else could repeat it.
- Choose software you can justify. Our comparison of SPSS, R and jamovi helps you decide.
Expected output: a single clean data file, a data dictionary and a source log.
What does a worked example look like? (illustrative)
This example is invented to show the process; it does not report real results. Suppose your question is whether schools in one English region with higher absence rates tend to have lower inspection outcomes. You download the school-level absence file from the DfE data catalogue and the latest inspection outcomes file from Ofsted, record their release dates, and join them on the school identifier. You check how many schools do not match and report that number. You group schools into absence bands, compare the proportion in each inspection outcome category, and test the association with a chi-square test if the cell counts allow it. In the discussion you point out that both measures describe schools, so you cannot say anything about individual pupils, and that a correlation does not show that absence causes a particular judgement. You also note that inspection frameworks and publication dates differ, so your comparison needs a clearly stated period.
Notice what the design does: it answers a question the data can actually support, and the limitations come directly from the unit of analysis and the source notes.
How do you join files from different publishers?
Joining DfE and Ofsted files usually means matching on a school identifier. Before you merge, check that the identifier has the same format in both files, that you are comparing the same time period and that you know what to do with schools that appear in one file but not the other. After the merge, count the matched and unmatched schools and report both numbers, because the unmatched ones may be closed, new or in a different category from the rest. If a large share fails to match, explain what you did about it rather than silently dropping it. Always keep an untouched copy of each original file, and save your merge steps as a short written procedure so that a marker could follow and repeat them.
How do you check the quality of public data?
- Read the methodology note. It explains definitions, coverage, revisions and known problems.
- Check the period. Academic years, financial years and inspection dates are not the same thing.
- Look for breaks in the series. A change in definition or in the collection method can make an apparent trend misleading.
- Check totals. If subgroup counts should add up to a total, test that they do, and note any rounding.
- Look at outliers. Very small schools or providers can produce extreme percentages, so decide how you will treat them and say why.
These checks take an afternoon and protect you from building an argument on a data quirk. They also give you material for your limitations section, which is exactly where markers expect to see an honest account of the data.
How do you cite and describe public data?
Treat a dataset like any other source. Name the publisher, give the title, the release or publication date, the URL and the date you accessed it, in your department’s referencing style. An illustrative Harvard pattern is: Department for Education (2026) Title of dataset [dataset]. Available at: URL (Accessed: 4 October 2026). In the text, mention the licence if your department asks for it, and state exactly which file and which release you used, because published files can be revised.
Do you need ethics approval for public data?
Often the risk is lower because the data are public and aggregated, but you should still follow your university’s process, because some departments require a form or a statement even for secondary data. If you plan to combine public data with your own survey or interviews, you need approval for the primary data collection. Our guide to ethics approval for an undergraduate dissertation explains what to check, and our overview of getting UK Data Service data covers registration if you want a survey dataset.
What are ten researchable education dissertation questions using public data?
- Is school absence in one region associated with inspection outcomes?
- How has the proportion of teachers leaving the profession changed in one region over a recent period?
- Do attainment gaps between groups of pupils differ between local authorities?
- How has the number of students in a named subject area changed across UK higher education providers?
- Do inspection outcomes differ between schools of different types in one region?
- How do suspension and exclusion rates vary across areas, and what school characteristics go with them?
- How has the gender balance in a subject area changed in higher education over time?
- Is the size of a school associated with its absence rate?
- How do school workforce characteristics differ between urban and rural areas?
- What does the data show about participation in a named subject at upper secondary level, and what questions does it leave open?
Each question names a unit, a source and a comparison. Check the latest release for each variable before you commit, because publications change names and coverage over time.
How do you present the results?
Lead with a table that gives the counts and proportions, then use one or two charts that show the comparison you care about. Label axes, give the source and the period under every table or figure, and keep the numbers you report consistent with your data file. Our guide to presenting the results of an education dissertation shows how to lay out tables and figures.
What mistakes do markers see?
- Writing questions about individual pupils from school-level data. This is the ecological fallacy.
- Assuming access to the National Pupil Database. It requires an eligible route and accreditation.
- Ignoring suppression and revisions. Report how you handled missing or rounded cells.
- Mixing years or definitions. Check the notes before you combine files.
- Not citing the release. State the title, date and access date for every file.
Tesify helps you structure the dissertation around your data source, so your question, methods and limitations stay aligned while you write every word yourself. Over 9,000 students have used it across more than 15,000 chapters. Plan your education dissertation with Tesify.
Education dissertation data FAQ
Where can I find data for an education dissertation in England?
Start with the Department for Education’s Explore education statistics service, which offers a data catalogue, a data table builder and an API, then add Ofsted inspection data and HESA open data as your question requires.
Can I use the National Pupil Database for my dissertation?
Only through an eligible route. Pupil-level data is shared by default through the ONS Secure Research Service, which requires ONS researcher accreditation, and direct access from DfE requires one of three exemptions, so it is not a realistic source for most student projects.
Is Ofsted inspection data public?
Yes. Ofsted publishes monthly management information on state-funded school inspection outcomes as ODS and CSV files on GOV.UK.
Can I reuse HESA data without paying?
HESA states that data published on its website is available under a Creative Commons Attribution 4.0 licence, so you can reuse it with attribution. Check the licence on the page you use.
Do I need ethics approval for public education data?
Follow your university’s process. Secondary public data is often lower risk, but some departments still require a form or a statement.
Does the data cover Scotland, Wales and Northern Ireland?
The DfE sources mostly describe England. Look for each nation’s own published statistics and state clearly which nation your data covers.
How do I cite a dataset?
Give the publisher, title, release date, URL and access date in your department’s referencing style, and say which file and release you used.
Can I combine a survey with public data?
Yes. Use the public data for context or comparison and your survey for the primary question, with ethics approval for the survey.
What is the biggest limitation of school-level data?
It describes schools, not individual pupils, so you cannot draw conclusions about individuals from it.