Question-by-question marking means marking the same question for every student before moving to the next question. Instead of completing Amelia’s whole paper and then Ben’s, you mark Question 1 across the class, followed by Question 2 across the class, and so on.
The method can make formal assessment marking more consistent because the same success criteria, mark scheme section and maximum mark remain active while you compare a complete set of responses.
Why mark one question at a time?
Every change of question has a mental setup cost. You locate a new part of the mark scheme, recall the available marks and adjust to a different type of answer. Paper-by-paper marking repeats that switch for every student.
Question-by-question marking groups similar decisions together. This can help with:
- Consistency: the standard applied to the first response is easier to compare with the thirtieth.
- Speed of navigation: the relevant answer and score control can remain in the same place.
- Misconception spotting: repeated patterns become visible while the full class set is fresh.
- Feedback reuse: a precise comment can be applied and adapted where the same issue recurs.
- Moderation: uncertain or boundary responses can be compared with nearby examples.
It does not remove the need for professional judgement. It gives that judgement a more stable context.
When question-by-question marking works well
The method is especially useful when an assessment has:
- discrete questions or subquestions;
- a clear mark scheme;
- a fixed maximum for each item;
- similar answer locations across scripts;
- a class-sized cohort;
- a need for question-level analysis.
Maths and science papers are obvious examples, but the approach also suits short-answer humanities questions, language exercises and rubric-scored sections of longer work.
When paper-by-paper marking may be better
Use the whole script when context across the response matters more than item-by-item comparison. An extended essay, portfolio or multi-stage investigation may need to be understood as one piece. Earlier reasoning can affect how later work is interpreted.
You can also combine the two methods. Mark structured sections question by question, then review the extended response paper by paper. A flexible digital marking workflow should support that decision instead of treating one order as a rule.
How to set up a question-first workflow
1. Confirm the assessment structure
List every scored question and subquestion in the same order as the paper. Check maximum marks and make sure the total matches the published assessment total. Attach the relevant mark scheme or rubric before marking begins.
2. Map or locate each answer
The method is fastest when the current answer can be shown directly. With paper scripts, that may mean using a template that records the answer area on each page. If the layout varies, the teacher still needs a quick way to expand to the full page and find continuation work.
3. Calibrate with a small sample
Read several responses before settling into the full run. Choose a likely high, middle and low response where possible. Use the mark scheme to clarify how you will treat common alternatives and record any local decision that the department should apply consistently.
Calibration should not turn into pre-marking the entire batch. Its purpose is to establish the standard and expose ambiguities early.
4. Mark the full question across the class
For each response:
- Read the evidence.
- Apply the mark scheme or rubric.
- Record the structured score.
- Add only useful feedback.
- Flag uncertainty rather than breaking the flow to solve every exception immediately.
- Move to the next student.
Keep the response, score and criteria visible together. If you must repeatedly open and close panels, the benefits of grouping the decisions start to disappear.
5. Review flagged and boundary responses
After the question is complete, return to flags as a group. Compare them with clearly awarded examples and check any departmental guidance. This is usually more consistent than interrupting the first pass whenever an unusual answer appears.
6. Record the class-level finding
Before moving on, note the misconception or strength that matters for teaching. This could be a common algebraic error, missing evaluation, weak evidence selection or a vocabulary gap. One concise observation is more useful than a dashboard full of undigested numbers.
How to keep question-by-question marking fair
The method makes comparison easy, but direct comparison needs discipline.
- Judge each response against the criteria, not against the previous student’s work.
- Hide student identity where blind marking is required.
- Avoid letting neat handwriting or answer length substitute for evidence.
- Use a clear state for unattempted answers rather than guessing intent.
- Revisit the first few scripts if the marking standard changed during calibration.
- Take a break between long question runs to prevent decision fatigue.
A rubric-based marking approach can make criteria more explicit, while the consistency guide covers calibration and review in more depth.
Turn marking into useful analysis
Question-first marking naturally creates question-level results. Once the scores are recorded, you can see which items had the lowest average, widest spread or highest rate of unattempted responses.
Interpret those patterns carefully. A low score might reveal a misconception, but it could also indicate an unclear question, insufficient teaching time or a mismatch between wording and prior practice. Use the script evidence alongside the numbers. The assessment results analysis guide provides a practical sequence.
Question-first marking in Nib
Nib is designed around the simple idea of showing one question across the whole class while keeping the original script available. Scores, feedback, rubrics, annotations and progress stay attached to the response, and uncertain work can be flagged for review.
For teachers moving a paper process onto iPad, start with the broader digital marking guide and trial the question-first method on one structured assessment.