Student evaluations should inform dialogue, not replace academic judgement
Updated Jun 29, 2026
student voicefeedbackStudent evaluations are easy to overuse because they arrive as neat scores and quotable comments. The harder question is what they should be allowed to decide. Maria Bengtsson's Teaching in Higher Education paper, "Reframing student feedback: from evaluation to dialogue", matters for universities building stronger student voice systems because it argues that feedback works best when it informs dialogue about learning rather than acting as a verdict on teaching quality.
Context and research question
In many universities, student evaluations of teaching sit between two aims that do not always fit comfortably together. One aim is to hear students' experience of a course. The other is to convert that feedback into evidence for quality assurance, review, or performance judgements. Bengtsson's paper asks what gets lost when those two aims are collapsed into one process, especially when student satisfaction starts to stand in for broader questions about learning, curriculum coherence, and academic standards.
This is not an empirical survey study or an experimental paper. It is a theory-led analysis of how student evaluations of teaching are used in Swedish higher education, and of what a better model might look like. Bengtsson draws on Biggs' theory of constructive alignment, Bernstein's pedagogic device, and Bourdieu's field theory to develop a dialogic account of curriculum development. For UK higher education teams, that makes the paper useful less as a benchmark and more as a disciplined way to think about where student evaluation evidence helps, where it misleads, and how it should be interpreted.
Key findings
The paper's starting point is that student feedback is valuable, but partial. Student evaluations capture lived experience of a course, which matters. Students can speak clearly about clarity, communication, organisation, inclusion, and whether teaching feels supportive. The problem begins when institutions treat that evidence as if it can directly settle the wider question of teaching quality or curriculum soundness on its own.
Bengtsson's warning is concise:
"overreliance on such evaluation risks elevating satisfaction over learning"
The second core argument is that academic judgement still matters because teachers can see things students often cannot see from one course encounter alone. Staff understand how a module fits into a longer programme, how expectations are staged across years, and why some tasks are designed to stretch students rather than simply please them. That does not make student feedback less important. It means the interpretation of that feedback needs to stay connected to curriculum intent and disciplinary judgement rather than being treated as a self-explanatory signal.
The paper therefore reframes evaluation as dialogue, not extraction. Instead of imagining that a survey produces a clean answer which staff then implement, Bengtsson argues for a more dialogic process in which student perspectives and educator expertise inform one another. That is a more realistic account of how teaching improvement happens. Comments and scores can surface issues, but the work of deciding what they mean still requires discussion, contextual knowledge, and professional responsibility.
A final strength of the paper is that it separates democratic participation from consumer-style verdicts. The Swedish context matters here because the discussion is framed around both democratic and scientific ideals. Students should have a meaningful voice in how education is experienced and improved. At the same time, that voice should not be turned into a simplistic market mechanism in which immediate satisfaction overrides longer-term educational purpose. For UK universities, that is a useful distinction when evaluation systems are under pressure to be both responsive and defensible.
Practical implications
The first implication for UK universities is to design evaluation workflows that lead to interpretation, not automatic judgement. If a module score drops or a cluster of comments turns negative, the next step should be a structured conversation about what the feedback is pointing to, what other evidence says, and what kind of action is appropriate. That approach fits what we already know from student survey benchmarking and triangulation: scores identify where to look, but they rarely explain the issue well enough on their own.
Second, institutions should separate what students can judge directly from what needs wider academic context. Students are usually well placed to comment on communication, respect, timeliness, clarity, and whether assessment expectations feel understandable. They are less well placed to judge on their own whether a curriculum is coherently sequenced across years or whether short-term discomfort serves a valid learning purpose. This is also why student evaluations improve when staff and students redesign them together is such a useful companion piece: better instruments ask clearer questions about the things students can genuinely see.
Third, universities should treat free-text comments as evidence for dialogue rather than ammunition for isolated reactions. A small set of sharp comments can distort judgement if it is read in isolation, especially when the same issue is not checked against repetition, cohort pattern, or other survey routes. A governed approach such as our NSS open-text analysis methodology helps institutions separate recurring concerns about organisation, support, assessment, or belonging from one-off frustration. The benefit is more credible follow-through and less risk of mistaking volume or immediacy for importance.
Finally, this paper is a reminder that student voice works best when institutions are clear about its purpose. If the goal is enhancement, comments should open discussion and guide redesign. If the goal is assurance, feedback still needs context, triangulation, and proportional use. In either case, universities are on firmer ground when they can show how student evidence was interpreted, not just collected. That is the difference between a survey process and a defensible feedback system.
FAQ
Q: How should a university discuss evaluation results with teaching staff after reading this paper?
A: Start with grouped themes rather than a raw bundle of comments, then ask three questions in order: what are students noticing, how widely does the pattern recur, and what wider course or programme context matters before action is taken? That process helps teams move from reaction to interpretation. It also reduces the risk that one difficult comment or one soft score shift is treated as decisive evidence.
Q: What should UK institutions keep in mind given that this paper is conceptual rather than empirical?
A: Its value lies in the framework, not in a new dataset. Bengtsson is not claiming to measure how often a problem occurs. She is showing why a familiar institutional habit, treating student evaluations as a direct verdict on teaching, can produce weak decisions. UK teams should read it as a strong interpretive lens for their own evaluation practice, then test the same questions against local surveys, committee discussions, and open comments.
Q: What broader implication does this have for student voice practice?
A: It suggests that student voice is strongest when it is built into a process of shared judgement rather than reduced to a satisfaction score. Students need routes to describe what they experienced, and institutions need methods for reading that evidence carefully, consistently, and in context. That makes feedback more useful for teaching teams and more credible for leaders who need to show how student evidence shaped decisions.
References
[Paper Source]: Maria Bengtsson "Reframing student feedback: from evaluation to dialogue" DOI: 10.1080/13562517.2026.2694024
Request a walkthrough
Book a free Student Voice Analytics demo
See all-comment coverage, sector benchmarks, and reporting designed for OfS quality and NSS requirements.
- All-comment coverage with HE-tuned taxonomy and sentiment.
- Versioned outputs with TEF-ready reporting.
- Benchmarks and BI-ready exports for boards and Senate.
UK-hosted · No public LLM APIs · Same-day turnaround