In the Postgraduate Research Experience Survey (PRES) 2025, run by Advance HE and reported in February 2026, 83 per cent of the 35,475 responding postgraduate researchers were satisfied with their experience overall; 89 per cent were positive about their supervision; but only 63 per cent were positive about their institution’s research culture — and among those who had considered leaving their programme, mental or emotional health was the most cited reason, at 20 per cent, ahead of financial concerns at 14 per cent.
PRES is the closest thing UK doctoral education has to a sector-wide experience dataset, and its 2025 wave is the one to quote this year. This page gives the figures in citable form and — because this survey is routinely misquoted — the scope notes that keep your citations honest.
The key figures
| Measure | Value | Source and year |
|---|---|---|
| Responses | 35,475 postgraduate researchers | PRES 2025 (Advance HE, Feb 2026) |
| Participating institutions | 93 — including 4 Australian institutions | PRES 2025 |
| Overall satisfaction | 83% | PRES 2025 |
| Positive about supervision | 89% | PRES 2025 |
| Positive about research culture | 63% | PRES 2025 |
| Considered leaving: mental or emotional health cited | 20% | PRES 2025 |
| Considered leaving: financial concerns cited | 14% | PRES 2025 |
What is PRES, exactly?
The Postgraduate Research Experience Survey is a sector benchmarking survey run by Advance HE, in which participating institutions invite their PGR students to report on supervision, resources, research culture, progression, wellbeing and skills development. Institutions use it to benchmark themselves against the aggregate; researchers and journalists use the aggregate as the sector’s experience barometer. It is an opt-in instrument at both levels — institutions choose to run it, students choose to answer — which is exactly where the caveats start.
The caveats that keep you honest
Three scope notes belong next to any quoted figure. The 2025 wave includes four Australian institutions among its 93, so it is not a purely UK snapshot — describe it as a survey of participating institutions, not “UK PhD students” simpliciter. The response rate was not published with the headline reporting, so you cannot say what fraction of the invited population answered, and self-selection cuts in unknowable directions — the burnt-out may answer disproportionately, or may be precisely the people with no evening to spare for a survey. And participation varies wave to wave, so cross-year comparisons are between overlapping but not identical sets of institutions. None of this makes PRES weak — 35,475 responses is a serious dataset — but each limits the sentences it can support, and the strategy of quoting the survey’s own numbers with its own scope attached is the one that survives review.
The supervision–culture gap is the finding
Read the three satisfaction figures together and the story is not “doctoral researchers are satisfied” or “doctoral education is in crisis” — it is a 26-point gap between supervision (89 per cent positive) and research culture (63 per cent). The relationship candidates have most control over, and institutions structure most tightly, scores highest; the diffuse institutional layer — community, seminar life, belonging, how researchers are treated between milestones — scores lowest. That gap is why sector attention (including the developing people-and-culture strand of research assessment) keeps moving toward environment, and it has a personal corollary: if your experience is good with your supervisors but thin everywhere else, you are not an outlier — you are the modal data point. The supervision side of that equation is the manageable one, and our guide to managing your supervisory team is about keeping it at the 89 per cent end.
The leaving question deserves its precise reading
The most quoted PRES numbers are the most misquoted. The 20 per cent and 14 per cent figures describe the reasons given by researchers who had considered leaving — mental or emotional health first, finances second — not the share of all PGR students in mental-health difficulty or financial hardship. Quote them as what they are: among doctoral researchers who thought about quitting, personal wellbeing outranks money as the stated driver. That ordering matters for policy (a stipend rise, tracked in our funding and stipends roundup, addresses the second reason, not the first) and for candidates: the sector’s own data says the pressure point is health, which makes treating wellbeing as a project-management variable — not an indulgence — the evidence-based position.
How institutions use PRES — and how you can use that
Inside a university, PRES results land on graduate school desks as benchmarked scores against the sector, broken down by faculty and question domain, and they feed action plans precisely because low outliers are visible to senior committees. That machinery is usable from below. If your graduate school runs the survey, answering it is one of the few costless levers a candidate has — free-text comments in these surveys are read, and a specific, fixable observation travels further than a general grievance. If you sit on a staff–student committee or a PGR forum, asking “what did our PRES results show and what changed” is a legitimate, precise question that administrators can answer and are often glad someone asked. And if you are choosing between institutions for a doctorate, whether a graduate school publishes or discusses its PRES results openly is itself a signal about the research culture the survey measures.
Using this data in your own writing
Attribute to the Postgraduate Research Experience Survey 2025, Advance HE (reported February 2026), and name the base for every percentage: all respondents for the satisfaction figures, considered-leavers for the reasons figures. Pair it with population data rather than substituting one for the other — PRES describes experience across ~35,000 respondents; the HESA enrolment statistics describe the full 121,175-person population — and resist deriving cross-survey trends unless you have checked that the question wording and participating institutions match. For institution-level claims, use your own university’s published PRES results if it releases them; the aggregate cannot speak for any single graduate school.
What the data suggests you actually do
Three evidence-aligned moves. Invest in the supervision relationship deliberately — it is the component with the highest satisfaction and the clearest management levers. Build culture locally where the institution’s is thin: writing groups, reading circles and peer review among candidates are research culture at the scale you control, and the 63 per cent figure is a licence to stop waiting for it to be provided. And treat steady progress as a wellbeing intervention: stalled writing and unclear milestones are exactly the chronic stressors a structured workflow removes. Tesify keeps the thesis visibly moving — chapters, drafts and bibliography structured in one place, 100% written by you — which is the controllable end of the mental-load problem the sector’s own survey keeps surfacing.
Frequently asked questions
What is PRES?
The Postgraduate Research Experience Survey — Advance HE’s recurring benchmarking survey of postgraduate researchers at participating institutions, covering supervision, resources, research culture, progression and wellbeing.
How many people answered PRES 2025?
35,475 postgraduate researchers across 93 participating institutions, four of them Australian. The response rate was not published with the headline reporting.
Are PhD students satisfied with their PhDs?
83 per cent of PRES 2025 respondents were satisfied overall — with the usual opt-in caveats. Satisfaction is highest for supervision (89 per cent) and much lower for research culture (63 per cent).
What proportion of PhD students have mental health problems?
PRES 2025 does not measure that. Its figure is narrower: among respondents who had considered leaving, 20 per cent cited mental or emotional health — the most common reason. Prevalence claims need clinical or dedicated wellbeing studies, not this survey.
Is supervision really the strongest part of the doctoral experience?
In the sector’s self-report data, consistently yes — 89 per cent positive in 2025. Individual experiences vary enormously, which is why the aggregate should never be used to dismiss a specific supervision problem.
Why is research culture rated so much lower?
The survey reports the gap rather than explaining it, but the pattern — strong dyadic supervision, weak surrounding community — recurs across waves and institutions, and is a stated driver of sector initiatives on research culture and environment.
Does PRES cover only the UK?
No — 2025’s institutions include four Australian participants. For strictly UK claims, either note the scope or use your institution’s own results.
Can I see PRES results for my university?
Institutions receive their own benchmarked results; some publish summaries internally or publicly. Ask your graduate school — and if you answered the survey, you have standing to ask what changed because of it.
How does PRES differ from the NSS?
The National Student Survey covers final-year undergraduates and is quasi-census; PRES covers postgraduate researchers at opt-in institutions. They measure different populations with different instruments and cannot be spliced.
Where is the full PRES 2025 report?
Advance HE publishes the aggregate reporting — the version to cite in academic or policy writing, and the place to check question wording before building any argument on a single percentage.