The Feedback Gap: From Assessment to Measurable Learning
Does generic feedback produce different outcomes from answer-specific feedback?
1. Does Assessment Lead to Better Learning?
A student receives 6/10.
The score records the outcome of an assessment, but does it explain what the student needs to do to improve?
- Which concept was misunderstood?
- Where did the reasoning break down?
- What essential information was missing?
- Did the student struggle with subject knowledge, interpretation, reasoning or expression?
These questions reveal a fundamental distinction between measuring performance and supporting improvement.
Assessment shows what a student has demonstrated at a particular point in time. Feedback can help interpret that performance, identify learning gaps and suggest appropriate next steps. However, providing feedback does not automatically result in better learning. Its effectiveness must be examined through subsequent student performance.
For digital assessment platforms such as ONETUZA, this raises an important research question: Does the type of feedback provided after an assessment influence measurable improvement in a student's subsequent performance?
To investigate this question, ONETUZA could conduct a controlled study comparing three approaches: marks alone, marks accompanied by generic comments, and marks accompanied by answer-specific feedback linked to assessment criteria.
By measuring performance before and after the feedback intervention, the study could examine whether different feedback approaches are associated with different learning outcomes.
The objective is not to assume that one approach is superior, but to establish what the evidence supports.
2. What Does Research Tell Us About Feedback?
Feedback is widely recognised as an important component of learning, but its effectiveness varies according to the information it provides and how learners use it.
The 2020 meta-analysis by Wisniewski, Zierer and Hattie, The Power of Feedback Revisited: A Meta-Analysis of Educational Feedback Research, examined findings from 435 studies involving more than 61,000 learners. The research reported an overall positive effect of feedback on learning, while highlighting variation in its effectiveness.
These findings do not establish that every form of detailed feedback is more effective than generic comments. Nor do they prove that rubric-linked, answer-specific feedback will consistently produce better outcomes across every subject, age group or learning context.
Instead, they reinforce an important distinction: the presence of feedback is not, by itself, evidence of improved learning.
The more meaningful question is whether feedback provides relevant, actionable information that helps students recognise their difficulties and address them.
For digital assessment, this distinction matters. A platform may generate marks, comments and performance summaries, but the educational value of these outputs must be evaluated through evidence of subsequent learning.
- Do students correct previous errors?
- Does their reasoning improve?
- Do their answers become more complete?
- Can they apply their understanding to a new question?
Answering these questions requires a systematic approach that connects assessment, feedback and subsequent performance.
3. A Mark Is an Outcome, Not an Explanation
Consider two students who both receive 6/10.
The first understands the underlying concept but omits an important explanation. The second misunderstands the concept itself. Although their scores are identical, their learning needs are different.
- Identified learning gap
- Potential next step
- Conceptual misunderstanding
- Clarify the underlying concept
- Incomplete reasoning
- Identify the missing logical step
- Inadequate coverage
- Address essential omitted points
- Misinterpretation of the question
- Practise identifying what the question requires
- Difficulty with expression
- Develop clearer, more precise explanations
A generic comment such as “Revise the chapter” may provide broad direction, but it does not necessarily explain the difficulty evident in an individual response.
More specific feedback could identify the relevant gap and suggest a practical next step. For example:
“Your answer identifies the main concept, but the explanation does not establish the relationship between cause and outcome. Review this connection and practise explaining it in two logical steps.”
This comment links the student's response to a particular weakness and an action that can be attempted.
Nevertheless, greater specificity should not automatically be equated with greater effectiveness. Whether this approach leads to better subsequent performance must be tested.
Organised assessment evidence may also help teachers identify recurring errors and distinguish between learning needs. Technology can support this process, while teachers retain responsibility for interpreting responses and deciding what intervention is appropriate.
4. Testing Three Feedback Approaches
ONETUZA could investigate the relationship between feedback and learning through a study involving three groups of students.
Group A: Marks alone
Students receive their assessment scores without additional written feedback. This provides a comparison condition for examining performance when no written guidance accompanies the mark.
Group B: Marks with generic feedback
Students receive their scores alongside broad comments such as “Improve your explanation,” “Revise the relevant concept” or “Include more supporting details.”
These comments offer general guidance but may not identify the precise gap in an individual response.
Group C: Marks with answer-specific feedback
Students receive their scores alongside feedback linked to their responses and the relevant assessment criteria. The feedback identifies an observed weakness and suggests an appropriate next step.
The three groups distinguish between receiving a score, receiving general guidance and receiving feedback connected more directly to assessment evidence.
All students would complete an initial assessment before receiving their assigned feedback. After a consistent learning interval, they would attempt a comparable assessment measuring the same learning objectives.
Their subsequent performance would then be compared to determine whether the feedback approaches are associated with different levels of improvement.
The central hypothesis is that answer-specific feedback may produce greater improvement.
5. Measuring Improvement Beyond the Score
The primary outcome should be the change in assessment performance between the initial and subsequent assessments.
Improvement = Subsequent assessment score − Initial assessment score
For example, a student who improves from 6/10 to 8/10 demonstrates an observed increase of two marks.
However, this change alone does not establish that feedback caused the improvement. Students may improve through revision, teaching support, practice or other influences. The results must therefore be compared across groups.
Marks also provide only one measure of learning. Additional outcomes could help explain the nature of any improvement:
- Conceptual accuracy: Has the student developed a more accurate understanding?
- Reasoning quality: Has the student corrected an incorrect or incomplete logical step?
- Answer completeness: Does the response address the essential assessment criteria?
- Transfer of learning: Can the student apply the concept to a different question?
- Error recurrence: Does the original misunderstanding persist?
- Retention: Is the improvement maintained in a later assessment?
These measures can help distinguish an increase in marks from a change in the quality of a student's understanding and responses.
For a meaningful comparison, the initial and subsequent assessments should measure the same learning objectives at comparable levels of difficulty. The questions should differ sufficiently to test application rather than simple recall of previous answers.
6. Ensuring a Credible Study
A well-designed study must distinguish the effects associated with feedback from other factors that influence performance.
The same marking rubric should be applied across all three groups. Where feasible, evaluators should score responses without knowing which feedback condition each student received, reducing the risk of scoring bias.
Students should receive comparable teaching, learning materials, practice opportunities and time between assessments. Initial attainment and relevant differences between students should also be considered during analysis.
Where feasible, random assignment to feedback groups would strengthen the study's ability to assess whether differences in outcomes are attributable to the feedback conditions. The required sample size should be established through an appropriate statistical power analysis.
Ethical safeguards are equally important. Student information must be protected, participants should receive appropriate information about the study, and necessary educational support should not be withheld. Relevant institutional approvals and consent should be obtained where required.
Finally, results should be reported objectively, including the magnitude and uncertainty of observed differences and the limitations of the study. Statistical significance alone is not enough to establish educational importance.
7. Interpreting the Findings Without Assumptions
The study may produce several outcomes, each offering useful insights.
If the answer-specific feedback group demonstrates greater improvement, the findings would support the potential value of evidence-linked feedback within the context studied.
If the generic and answer-specific feedback groups perform similarly, broad comments may be sufficient for certain learning objectives or student groups. Further investigation could identify where greater specificity matters.
If no meaningful differences emerge, the results would challenge the assumption that the tested feedback approaches necessarily produce different outcomes. The study design, feedback quality and assessment measures may warrant further examination.
The effects may also differ according to the nature of the learning gap. Conceptual misunderstandings may respond differently to feedback than incomplete explanations or errors in reasoning.
In every case, the conclusions must reflect the evidence collected. The purpose is not to validate a preferred approach, but to understand which feedback methods appear useful, for whom, and under what conditions.
8. The ONETUZA Opportunity: Connecting Evaluation with Learning
The proposed study offers ONETUZA an opportunity to examine assessment as a connected learning process rather than an isolated scoring event.
The framework is:
Student Response → Assessment Evidence → Criteria → Identified Gap → Feedback → Subsequent Attempt → Measured Improvement
Each stage has a distinct purpose. The student's response provides evidence of current understanding. Assessment criteria establish the basis for evaluation. The identified gap connects observed performance with a possible area for improvement. Feedback communicates actionable information, while the subsequent attempt provides new evidence of performance.
ONETUZA could use this framework to investigate whether connecting feedback more explicitly to assessment evidence is associated with measurable improvements in subsequent work.
The emphasis is not on producing more comments or assuming that detailed feedback is necessarily better. It is on examining whether feedback is relevant, actionable and followed by demonstrable improvement.
Crucially, technology should support rather than replace teachers' professional judgement. Teachers remain central to interpreting responses, identifying learning needs and determining appropriate interventions. A digital platform can help organise assessment evidence and make the relationship between feedback and subsequent performance easier to examine.
9. Closing the Loop: From Performance to Progress
Traditional assessment often concludes when a score is awarded. Feedback may follow, but the connection between the original evaluation and subsequent learning is not always systematically examined.
A more complete approach asks what happens next.
Did the student correct the original misunderstanding? Did the explanation become more coherent? Could the student apply the concept to a new question? Or did the same difficulty persist?
These questions shift attention from feedback delivery to feedback effectiveness.
For ONETUZA, the opportunity lies in investigating whether connecting assessment evidence with targeted feedback can support measurable improvements in subsequent performance. The findings could inform future assessment practices and help identify where different feedback approaches may be useful.
The outcome must remain evidence-led. Improvement after feedback does not, by itself, establish causation; a carefully designed comparison is needed to estimate its contribution.
A score records performance. Feedback identifies a possible path forward. The next attempt provides evidence of progress.
The real question is not simply whether feedback was given, but whether there is measurable evidence of improved learning.