NSS open-text research brief · 2026 edition
What Computer Science students said in NSS 2026
Teaching Staff is the most frequently mentioned reportable topic in 2026.
01 · The 2026 answer
What changed from 2025?
The 2026 analysis covers 1,494 classified comments in Computer Science.
Teaching Staff is the leading reportable topic in this brief. Its mention rate changed by −1.0 percentage points and its sentiment index by −6.0 from 2025.
Teaching Staff
02 · Findings
Where sentiment differs from the sector
Topics within this subject are shown only when at least 20 comments support the cut.
Above-sector sentiment
-
Personal Development
n=68 · sentiment +61.3 · +0.6 points vs sector · 4.6% mention rate
-
Campus, City and Location
n=53 · sentiment +55.7 · +20.3 points vs sector · 3.5% mention rate
-
Library
n=43 · sentiment +55.6 · +9.8 points vs sector · 2.9% mention rate
Below-sector sentiment
-
Communication About Course and Teaching
n=54 · sentiment −35.4 · −8.4 points vs sector · 3.6% mention rate
-
Feedback
n=243 · sentiment −25.4 · −12.9 points vs sector · 16.3% mention rate
-
Student Voice
n=40 · sentiment −18.4 · −14.8 points vs sector · 2.7% mention rate
03 · Comparisons
Topics compared with the sector
Subject and sector figures are calculated on a like-for-like basis using the same deterministic supervised learning approach.
| Topic | n | Mention rate | Sentiment | Vs sector |
|---|---|---|---|---|
| Teaching Staff | 439 | 29.4% | +22.6 | −19.3 |
| Type and Breadth of Course Content | 426 | 28.5% | +20.1 | −5.8 |
| Delivery of Teaching | 394 | 26.4% | +7.9 | −15.0 |
| Feedback | 243 | 16.3% | −25.4 | −12.9 |
| Student Support | 229 | 15.3% | +35.1 | +0.6 |
| Organisation and Management of Course | 185 | 12.4% | −5.6 | +2.6 |
| Assessment Methods | 176 | 11.8% | −11.5 | +5.5 |
| Learning Resources | 142 | 9.5% | +22.2 | −1.3 |
| Module Choice and Variety | 120 | 8.0% | +21.8 | +0.8 |
| Student Life | 111 | 7.4% | +39.0 | −4.3 |
04 · Time series
Current questionnaire period, 2023–2026
The 2023 NSS questionnaire redesign creates a comparability break. We show earlier years separately as context rather than drawing a trend through 2022–2023.
| Year | Comments |
|---|---|
| 2023 | 1,191 |
| 2024 | 1,719 |
| 2025 | 1,679 |
| 2026 | 1,494 |
Show historical context, 2018–2022
All years were analysed with the same deterministic supervised learning approach, but the survey instrument differs from the current questionnaire.
| Year | Comments |
|---|---|
| 2018 | 1,799 |
| 2019 | 2,027 |
| 2020 | 1,791 |
| 2021 | 1,173 |
| 2022 | 1,525 |
05 · Action
Three evidence-linked actions
Use the findings to choose a local test, then check the same topic and cohort again rather than treating a sector pattern as a diagnosis of one provider.
-
1
Turn criteria into shared judgements
Use plain-language criteria, annotated work at different standards and marker calibration before high-volume assessment begins.
Evidence: Marking Criteria has a 2026 sentiment index of −41.3 from n=105 comments.
-
2
Manage workload at programme level
Map expected effort and deadlines across modules, smooth avoidable peaks and remove duplicated tasks that do not add learning value.
Evidence: Workload has a 2026 sentiment index of −38.4 from n=34 comments.
-
3
Create one source of course truth
Use one maintained location for timetables, assessment information and course changes, with named owners and a short change log.
Evidence: Communication About Course and Teaching has a 2026 sentiment index of −35.4 from n=54 comments.
06 · Method and limits
How to read this evidence
How topics are identified
Deterministic supervised learning models identify topics in each sentence. A comment counts once in every topic it mentions; mention rate is the share of comments included in the analysis for the same population, so topic rates do not sum to 100%.
Sentiment index
The index summarises the balance of positive and negative language from −100 to +100. Scores are averaged within each comment first, so longer comments do not carry more weight.
When results are shown
Pages require at least 100 comments and three reportable topics or subject cuts. Displayed cuts require n≥20; 2026-versus-2025 claims require n≥30 in both years.
Scope
This is authorised aggregate analysis of OfS NSS national undergraduate open-text comments. In 2026, 40,822 of 43,870 source comments were classified (93.1%); mention-rate denominators exclude unclassified comments.
07 · Reuse
Cite this page
Student Voice research team (2026). “Computer Science student feedback: NSS open-text insights, 2026.” Reviewed by Dr Stuart Grey. Student Voice AI. https://www.studentvoice.ai/cah3/computer-science/