Assessment & Feedback
NSS Assessment and Feedback Scores:
Why They Stall and What Actually Moves
Them
NSS Assessment and Feedback Scores:
Why They Stall and What Actually
Moves Them
Most institutions have already tried the obvious fixes: tighter turnaround
policies, updated rubrics, staff briefings. Yet Assessment and Feedback
remains the lowest scoring NSS category, year after year. Here is what
the research actually says about why scores stall and what closes the
gap.
Eduface
·
8 min read
·
Written for DVCs & PVCs
Assessment and Feedback. Year after year, it sits near the bottom of your NSS results.
You have read the reports, attended the sector briefings, and tightened your turnaround
policies. The scores nudge up by a point, then slide back again. The problem is unlikely
to be effort. It is that most interventions target the visible symptoms rubric updates,
policy deadlines without addressing the root cause: the consistency and depth of
feedback that students actually receive on every piece of work, across every module,
throughout the year. NSS assessment and feedback scores stall for structural reasons,
and structural problems require structural solutions.
Why are NSS assessment and feedback scores so difficult to improve?
NSS assessment and feedback scores are difficult to shift because the problem is
structural, not behavioural. Students want feedback that is specific, timely, and
actionable on every submission. Lecturer workloads make this difficult to sustain at
scale. Institutions that successfully move their scores combine clearer marking
criteria, faster turnaround, and a mechanism for giving every student detailed,
personalised comments on every piece of work, regardless of cohort size.
What do the NSS assessment and feedback questions actually
measure?
The NSS Assessment and Feedback category contains five questions. They ask
whether marking criteria were made clear in advance, whether feedback was detailed
enough to be useful, whether it arrived on time, whether it helped students understand
their performance, and whether marking has been fair. These questions do not
measure administrative process. They measure whether students felt their work was
taken seriously.
This distinction matters in practice. An institution can comply with a 15-working-day
turnaround policy and still score poorly if the feedback students receive is brief or
generic. Speed matters, but it addresses only one of the five questions. The remaining
four depend on depth, specificity, and perceived fairness all of which are far harder
to improve through policy alone.
The 5 NSS Assessment & Feedback Questions
NSS
A&F Score
Criteria clear
in advance
Feedback
on time
Helped me
improve
Understood
performance
Marking
was fair
Policy changes typically address one or two of these five questions. Students score all five and the overall
category score reflects the weakest link.
NSS question focus
What institutions
typically fix
What actually moves the score
Feedback received on time
Turnaround policies and
monitoring
Structured delivery workflows with
escalation alerts
Feedback helped me
understand my
performance
Rubric clarity and
standardisation
Specific, criterion-referenced
written comments per submission
Feedback helped me
improve my work
Staff training and guidance
sessions
Actionable, personalised
comments tied directly to the
student's work
Marking criteria clear in
advance
Publishing rubrics online
Pre-submission briefs with worked
examples and sample feedback
Marking has been fair
Internal moderation
processes
Consistent depth and structure of
feedback across the whole cohort
Why do NSS assessment and feedback scores resist improvement
despite sustained effort?
The core tension is scale. A lecturer with 90 students and three assessed pieces per
semester faces 270 marking events. Writing meaningful, specific feedback on each
takes time that most timetables do not protect. Under pressure, feedback becomes
shorter. Shorter feedback is quicker to write, but research consistently shows it is
harder for students to act on.
1
And because NSS scores reflect experience
accumulated across the whole year, a single improved module rarely shifts the
institution wide aggregate.
There is also a consistency problem. Two lecturers applying the same rubric will
produce feedback of different depth and specificity in practice. Students compare
notes. When a peer receives more detailed comments on an equivalent piece of work,
it registers as unfair and that perception feeds directly into how students answer
NSS questions about fairness and usefulness.
2
The arithmetic of feedback at scale
90
students
×
3
assessments
=
270
marking events
per semester
At 15 min of written feedback per submission = over 67 hours of writing per semester.
Most quality failures in assessment feedback are not attitudinal, they are arithmetical. Without structural
support, consistency is almost impossible to maintain at this volume.
What does the evidence say about what actually moves NSS
assessment and feedback scores?
Three factors appear consistently in institutions that have improved their Assessment
and Feedback results over time:
1
Feedback specificity. Students respond to feedback that references their actual
submission, not a general description of the mark band. Hattie and Timperley's
widely cited meta-analysis of over 500 studies found that the most effective
feedback is task-specific, process-focused, and oriented towards improvement
rather than evaluation.
1
Comments that say "your argument in section two lacks
a supporting citation" are rated significantly more useful than "further evidence
needed."
2
Turnaround within 15 working days. This is now the sector standard in the UK.
Missing it reliably suppresses scores. Meeting it does not, on its own, improve
them because turnaround corresponds to only one of the five NSS questions.
3
Consistency across the cohort. Nicol and Macfarlane-Dick identify consistency in the feedback experience as a prerequisite for students developing self regulatory learning skills. Students need to be able to predict and trust the feedback process before they can engage with it productively.
2
Inconsistency within a cohort does not just lower NSS scores,
it undermines learning.
How are AI assessment tools changing the NSS assessment and
feedback picture?
AI-powered assessment platforms make it possible to give every student a detailed,
criterion-referenced response on every submission, regardless of cohort size. The
lecturer sets the rubric and quality thresholds. The system generates structured
feedback aligned to those criteria. The lecturer reviews and approves before anything
reaches students. The result is specific, consistent feedback that arrives on time. Staff
time is redirected towards the submissions that need the most careful human attention.
This is the human-in-the-loop model required by the EU AI Act. The lecturer remains
the assessor throughout. The AI addresses the scale and consistency problem that has
made meaningful feedback so difficult to sustain at cohort level, and that policy alone
has consistently failed to solve.
How AI-assisted assessment works at Eduface
Lecturer defines
rubric and criteria
Eduface assesses
every submission
Lecturer reviews
and approves
Human-in-the-loop
Student receives
specific feedback
At every stage, the lecturer retains control over what is released to students. The AI handles volume and
consistency; the lecturer handles judgement and accountability.
95%
In UK pilot programmes, Eduface's automated
assessments aligned with lecturer marks in 95 per cent of
cases. This is not a replacement for professional
judgement, it is a replication of it at cohort scale.
Eduface UK pilot data, 2023–2024. Pilot partners include Bath Spa University
and De Haagse Hogeschool.
Pilot institutions have observed that students engage more actively with feedback
when it is specific to their own work. Active engagement with feedback is precisely
what shifts NSS scores on the questions that matter most: whether feedback helped
students understand their performance, and whether it helped them improve.
Frequently asked questions
Q
Does faster feedback turnaround automatically improve NSS assessment and
feedback scores?
Not automatically. Turnaround time corresponds to one of the five NSS questions in
this category. Institutions that meet the 15 working-day standard but still score poorly
are typically underperforming on specificity and perceived fairness, not speed.
Meeting the turnaround threshold is a necessary condition for a competitive score, not
a sufficient one.
Q
Can AI-generated feedback be accurate enough for formal assessment?
In Eduface's UK pilots, AI-generated assessments aligned with lecturer marks in 95
per cent of cases. The platform operates on a human-in-the-loop basis: all feedback
is reviewed and approved by the relevant lecturer before it reaches students. The
accuracy standard is the same as for any structured marking support tool, and the
final decision always rests with the lecturer.
Q
Will students know if their feedback was produced with AI assistance?
That is an institutional decision, and transparency is generally advisable. Evidence
from pilot programmes suggests that students care far more about whether feedback
is specific and useful than about the mechanism behind it. Generic feedback written
by a human is consistently rated as less useful than specific, criterion-referenced
feedback produced with AI assistance and approved by a lecturer.
Q
Does Eduface integrate with our existing VLE?
Yes. Eduface integrates with Blackboard, Brightspace, Moodle, and Canvas.
Submissions and feedback flow through the institution's existing LMS, so there is no
new interface for students or lecturers to learn.
Q
Is Eduface compliant with GDPR and the EU AI Act?
Yes. Eduface runs on proprietary GPU infrastructure in the Netherlands and does not
pass student data to external AI providers such as OpenAI. It is fully compliant with the
EU AI Act's human in the loop principle and is an approved supplier on the
Jisc/CHEST framework in the UK and the HEAnet framework in Ireland.
NSS assessment and feedback scores are hard to move because the problem is
structural: lecturers cannot consistently deliver specific, detailed feedback at scale
without support. Institutions that are closing the gap combine clearer criteria with AI-
assisted tools that give every student a substantive response on every submission. If
your scores have stalled despite sustained policy effort, the missing intervention is
almost certainly one that addresses consistency and depth rather than speed.
See what AI-assisted assessment looks like with your
own assignments
Create a free lecturer account and test Eduface with a real
cohort, no contract and no time limit.
Create free account
Request a demo
References
Hattie, J., & Timperley, H. (2007). The power of feedback. Review of Educational Research, 77(1), 81–
112.
Nicol, D. J., & Macfarlane-Dick, D. (2006). Formative assessment and self-regulated learning.
Studies
in Higher Education, 31
(2), 199–218.
Black, P., & Wiliam, D. (1998). Assessment and classroom learning. Assessment in Education, 5(1), 7–
74.
Assessment & Feedback
NSS Assessment and Feedback
Scores: Why They Stall and
What Actually Moves Them
Most institutions have already tried the obvious
fixes. Yet Assessment and Feedback remains the
lowest-scoring NSS category, year after year.
Here is what the research says about why scores
stall and what closes the gap.
Eduface
·
8 min read
·
DVCs & PVCs
Assessment and Feedback. Year after year, it
sits near the bottom of your NSS results. You
have read the reports, tightened your
turnaround policies. The scores nudge up by a
point, then slide back again. The problem is
unlikely to be effort. It is that most interventions
target the visible symptoms without addressing
the root cause: the consistency and depth of
feedback students actually receive on every
piece of work, across every module. NSS
assessment and feedback scores stall for
structural reasons, and structural problems
require structural solutions.
Why are NSS assessment and feedback
scores so difficult to improve?
NSS assessment and feedback scores are
difficult to shift because the problem is
structural, not behavioural. Students want
feedback that is specific, timely, and
actionable on every submission. Lecturer
workloads make this difficult to sustain at
scale. Institutions that successfully move
their scores combine clearer marking
criteria, faster turnaround, and a mechanism
for giving every student detailed,
personalised comments regardless of
cohort size.
What do the NSS assessment and
feedback questions actually measure?
The NSS Assessment and Feedback category
contains five questions. They ask whether
marking criteria were clear in advance, whether
feedback was detailed enough to be useful,
whether it arrived on time, whether it helped
students understand their performance, and
whether marking has been fair. These
questions do not measure administrative
process. They measure whether students felt
their work was taken seriously.
An institution can comply with a 15-working-
day turnaround policy and still score poorly if
the feedback students receive is brief or
generic. Speed matters, but it addresses only
one of the five questions. The remaining four
depend on depth, specificity, and perceived
fairness all of which are far harder to
improve through policy alone.
The 5 NSS Assessment & Feedback Questions
NSS
A&F Score
Criteria clear
in advance
Feedback
on time
Helped me
improve
Understood
performance
Marking
was fair
Policy changes typically address one or two of these five
questions. Students score all five.
NSS question focus
Typical fix
What actually moves the score
Feedback on time
Turnaround policies
Structured delivery workflows
with alerts
Feedback helped
understand performance
Rubric clarity
Specific, criterion-referenced
comments
Feedback helped improve
work
Staff training
Personalised comments tied to
the student's work
Marking criteria clear in
advance
Publishing rubrics
Pre-submission briefs with
worked examples
Marking has been fair
Internal moderation
Consistent depth across the
whole cohort
Scroll table to see all columns →
Why do NSS scores resist improvement
despite sustained effort?
The core tension is scale. A lecturer with 90
students and three assessed pieces per
semester faces 270 marking events. Writing
meaningful, specific feedback on each takes
time that most timetables do not protect. Under
pressure, feedback becomes shorter and
shorter feedback is harder for students to act
on.
1
There is also a consistency problem. Two
lecturers applying the same rubric will produce
feedback of different depth and specificity.
When a peer receives more detailed comments
on an equivalent piece of work, it registers as
unfair and that perception feeds directly into
how students answer NSS questions about
fairness and usefulness.
2
The arithmetic of feedback at scale
90
students
×
3
assessments
=
270
marking events
per semester
At 15 min per submission = over 67 hours of writing per semester.
Most quality failures in assessment feedback are arithmetical,
not attitudinal.
What does the evidence say about what
actually moves NSS scores?
Three factors appear consistently in institutions
that have improved their Assessment and
Feedback results:
1
Feedback specificity. Students respond
to feedback that references their actual
submission. Hattie and Timperley's meta analysis of over 500 studies found that
the most effective feedback is task-
specific, process-focused, and oriented
towards improvement.
1
2
Turnaround within 15 working days. This
is the sector standard. Missing it reliably
suppresses scores. Meeting it does not,
on its own, improve them turnaround
corresponds to only one of the five NSS
questions.
3
Consistency across the cohort. Nicol
and Macfarlane-Dick identify consistency in the feedback experience as a prerequisite for students developing self regulatory learning skills.
2
Inconsistency does not just lower NSS scores it undermines learning.
How are AI assessment tools changing
the picture?
AI-powered assessment platforms make it
possible to give every student a detailed,
criterion referenced response on every
submission, regardless of cohort size. The
lecturer sets the rubric and quality thresholds.
The system generates structured feedback
aligned to those criteria. The lecturer reviews
and approves before anything reaches
students.
This is the human-in-the-loop model required by the EU AI Act. The lecturer remains the assessor throughout. The AI addresses the scale and consistency problem that policy alone has consistently failed to solve.
How AI-assisted assessment works at Eduface
Lecturer defines
rubric and criteria
Eduface assesses
every submission
Lecturer reviews
and approves
Human-in-the-loop
Student receives
specific feedback
The AI handles volume and consistency; the lecturer handles
judgement and accountability.
95%
In UK pilot programmes, Eduface's
automated assessments aligned with
lecturer marks in 95 per cent of cases — a replication of professional judgement at
cohort scale.
Eduface UK pilot data, 2023–2024. Bath Spa University and
De Haagse Hogeschool.
Pilot institutions have observed that students
engage more actively with feedback when it is
specific to their own work and that
engagement is precisely what shifts NSS
scores on the questions that matter most:
whether feedback helped students understand
their performance, and whether it helped them
improve.
Frequently asked questions
Q
Does faster feedback turnaround
automatically improve NSS scores?
Not automatically. Turnaround time
corresponds to one of the five NSS
questions. Institutions that meet the 15-
working-day standard but still score poorly
are typically underperforming on
specificity and perceived fairness, not
speed.
Q
Can AI-generated feedback be
accurate enough for formal
assessment?
In Eduface's UK pilots, AI-generated
assessments aligned with lecturer marks in
95 per cent of cases. All feedback is
reviewed and approved by the relevant
lecturer before it reaches students.
Q
Will students know if their feedback
was produced with AI assistance?
That is an institutional decision. Evidence
from pilot programmes suggests students
care far more about whether feedback is
specific and useful than about the
mechanism behind it. Specific feedback
approved by a lecturer is consistently rated
more useful than vague human-written
comments.
Q
Does Eduface integrate with our
existing VLE?
Yes. Eduface integrates with Blackboard,
Brightspace, Moodle, and Canvas.
Submissions and feedback flow through
the institution's existing LMS, so there is no
new interface for students or lecturers to
learn.
Q
Is Eduface compliant with GDPR and
the EU AI Act?
Yes. Eduface runs on proprietary GPU
infrastructure in the Netherlands and does
not pass student data to external AI
providers. It is fully compliant with the EU
AI Act's human-in-the-loop principle and is
an approved supplier on Jisc/CHEST (UK)
and HEAnet (Ireland).
NSS assessment and feedback scores are hard
to move because the problem is structural:
lecturers cannot consistently deliver specific,
detailed feedback at scale without support.
Institutions that are closing the gap combine
clearer criteria with AI-assisted tools that give
every student a substantive response on every
submission. If your scores have stalled despite
sustained policy effort, the missing intervention
is almost certainly one that addresses
consistency and depth rather than speed.
See what AI-assisted
assessment looks like with your
own assignments
Create a free lecturer account and test
Eduface with a real cohort, no contract
and no time limit.
Create free account
Request a demo
References
Hattie, J., & Timperley, H. (2007). The power of
feedback. Review of Educational Research, 77(1), 81–
112.
Nicol, D. J., & Macfarlane-Dick, D. (2006). Formative
assessment and self-regulated learning.
Studies in
Higher Education, 31
(2), 199–218.
Black, P., & Wiliam, D. (1998). Assessment and
classroom learning. Assessment in Education, 5(1), 7–
74.