Student Voice Analytics: student survey comment analysis

From comments to priorities

Student Voice Analytics helps university survey and student experience teams organise open-text feedback, examine differences between groups and prepare reports for people who need to act. It works with exported comments alongside your existing survey collection process.

Start with one survey and a clear decision: which assessment problems need investigation, where support is working, or what to discuss with programme teams. Agree the population, reporting groups and delivery date before analysis begins.

Analysis designed for higher education

Advance HE’s PRES 2025 sector report (page 6) identifies evasys and Student Voice AI as providers of optional analysis of the survey’s two main open-comment questions, delivered directly to participating institutions on behalf of Advance HE.

• Relevant categories: organise comments around teaching, assessment, feedback, support, resources and other aspects of university life.
• Context within a comment: a student can praise teaching and criticise feedback in the same response; topics and sentiment need to preserve that distinction.
• Coverage you can inspect: report how many comments were supplied, retained and classified, with reasons for exclusions. Processing all supplied comments does not eliminate non-response bias or classification error.
• Useful comparisons: examine cohorts and years with comparable questions, definitions and sufficient data. Use sector context where an appropriate reference is available.

How an analysis project works

• Agree the question. Choose the survey, years, reporting groups, outputs and deadline. Turnaround is agreed for your data and reporting needs.
• Prepare the export. Supply comment text with survey, year and necessary group fields through the agreed transfer route. Avoid collecting identifiers that the analysis does not need.
• Analyse and review. Categorise the comments, inspect coverage and review examples and ambiguous cases before interpreting patterns.
• Report and follow up. Discuss findings with local teams, agree action owners and use a comparable analysis in later cycles to examine change.

What to ask to see in a demo

• A summary of the main positive and negative themes, with denominators and limitations.
• A category-level report and permitted school, programme or cohort comparisons.
• Examples showing how source comments support a finding, with disclosure controls appropriate to the audience.
• Spreadsheet or BI export options, field definitions and a description of the analysis version.

The public NSS research briefs show our approach to aggregate reporting. They are openly available sector research; your institution’s private analysis is scoped separately.

Method and governance

Our public NSS release uses a fixed supervised-learning method and publishes classification coverage, sample thresholds and comparison rules. Read the methodology to understand what the measures mean and where they are limited.

For an institutional project, agree data access, retention, processing locations, permitted reporting and the treatment of small groups before transfer. Record model and category versions when comparing survey cycles. A repeatable method supports scrutiny; it does not itself establish accuracy or explain a causal effect.

Choose the right workflow

For a close reading of a small set of interviews or comments, researcher-led coding may be the right approach. For recurring institution-wide surveys, compare the quality of the categorisation, reporting effort and governance as well as processing speed. Use our NSS comment analysis buyer’s guide or Explorance MLY comparison to structure an evaluation.

Survey-specific starting points

• Module evaluation comment analysis
• PTES and PRES comment analysis
• NSS comment sector benchmarking

Discuss a survey

Tell us which survey you use, roughly how many comments you have and the decision or deadline the analysis needs to support. We can discuss suitable outputs and show examples.

Common questions

Which student surveys can we discuss?

NSS, PTES, PRES, UKES, module evaluations and other student feedback surveys. Categories, comparison groups and deliverables are agreed for each project.

Does analysing every comment remove bias?

No. Full-corpus processing avoids selecting only a subset of supplied comments, but non-response, exclusions and classification errors still need to be considered.

Is a next-day turnaround guaranteed?

Turnaround is agreed for your data, review requirements and reporting deadline before work starts.

Can we explore the research before contacting you?

Yes. The public NSS research hub and methodology are freely accessible and do not require a demo or form submission.