Build an assessment blueprint by mapping the taught curriculum to the evidence each task must elicit. Allocate prompts deliberately, check access and timing, identify the reviewer and decide what the results can support before writing the paper. A blueprint makes those choices visible. It doesn't make a teacher-made assessment a validated measure.
This annotated example is for teachers, assessment leads, SENCos and school leaders in England. It specifies one small assessment across several taught ideas. For a single in-lesson decision, use the separate classroom hinge routine.
Fix the claim and teaching boundary
Start with the decision: which taught ideas need further explanation or experience before the class moves on? Write down the content included, the content omitted and the responses that would help the teacher decide. A paper assembled from convenient questions can quietly become an assessment of whatever was easiest to photocopy.
NFER's guidance on test validity connects the intended use of results with appropriate curriculum sampling and removal of irrelevant demand. A short check can support a narrow teaching decision while leaving much of the subject unassessed. It cannot support a whole-subject judgement simply because its total fits in a markbook.
The Standards and Testing Agency's test development handbook describes specifications, content and cognitive domains, trials and expert review for national tests. That is a different development process from classroom planning. Borrow the discipline of specifying coverage; don't claim the resulting classroom paper has undergone those trials or acquired national-test comparability.
Annotated blueprint: Year 4 electricity
Everything in this planning record is synthetic, including the staff, dates and evidence. It describes no pupil, school or customer. Ms Lewis, the fictional class teacher, owns the interpretation. Mr Owen, the fictional science lead, reviews the design with her.
Purpose and boundary: plan follow-up teaching about complete circuits, switches and conductors after the class's recorded teaching sequence of 2 to 10 November 2026. The Year 4 science programme of study includes these ideas. This sample omits identifying appliances and direct evidence of constructing a circuit. It therefore cannot establish coverage of the whole electricity programme. Use familiar pictorial circuits; conventional circuit symbols are introduced in Year 6.
Record header, version 1: keep the curriculum plan reference, dated lesson log, task sheet, response guidance and blueprint together. Ms Lewis checks teaching exposure and support on 12 November. The provisional assessment window is 13 November, with a design and response review on 16 November and a follow-up sample on 20 November. These are invented local planning dates, not national deadlines. Move the window if teaching or access isn't ready.
| Taught content and allocation | Intended demand and prompt | Access and evidence to retain | Interpretation limit |
|---|---|---|---|
| Complete circuits: three prompts | Name familiar components; identify a complete loop in two pictorial circuits; explain why a pictured gap prevents the lamp lighting. | Keep pictures clear and response instructions short. Accept pointing and an oral explanation. Retain the actual response and any adult cue. | Naming parts doesn't establish understanding of the loop. Pictures don't demonstrate independent construction with equipment. |
| Switches: two prompts | Interpret a pictured open switch; explain how closing it changes the loop in an otherwise complete, working circuit. | Use the familiar picture format. Record whether the adult points to the gap or supplies an explanation. | A correct prediction without an explanation leaves the reasoning uncertain. A prompt that reveals the gap changes the evidence. |
| Conductors: three prompts | Identify a familiar conductor; use taught observations to choose a material for a gap; explain the choice using those observations. | Use materials already studied. Retain the source observation, conditions and response. Any demonstration needs checked classroom equipment and school safety procedures. | A material result depends on a working circuit and sound connections. One example cannot establish that every material of its appearance conducts. |
Read the allocation: eight prompts cover three selected content areas. Naming, interpreting and explaining describe what pupils are asked to do; they aren't validated difficulty tiers. A familiar explanation may be easier than an unfamiliar naming task. This allocation is an editorial choice for the stated purpose, not a recommended national weighting or a percentage-of-mastery calculation.
Apply this review contract to every row: Ms Lewis records the observer, source and date, task version, curriculum exposure, conditions, usual support and adjustments, adult contribution and missing responses. All interpretations remain provisional until the 16 November review. Further work, a clarified response or an overlooked cue can change them. Pupils can explain or correct their account with her; parents can contact her with context; the SENCo or science lead can question the support or interpretation. Keep corrections with their author, date and reason alongside the initial record.
Run the blueprint through a preflight
- Check coverage against actual teaching. Read the lesson log beside each row. Remove untaught material or provide the missing teaching before collecting evidence.
- Try the questions and response guidance. Ask a colleague what each prompt elicits. Identify plausible alternative answers, ambiguous pictures and wording that supplies the answer.
- Check access without losing the question. Retain established communication support and record its use. If handwriting isn't the intended evidence, a scribed response may preserve it; check the meaning with the pupil and distinguish adult wording.
- Test the proposed use. Decide which response would prompt reteaching, another observation or a question for the pupil. If a total cannot explain that decision, preserve the content-level evidence.
Ofqual's Section D conditions and guidance concern awarding organisations and regulated qualifications. They distinguish the intended knowledge or skill from irrelevant accessibility barriers. Here, that distinction prompts a classroom question: is a long reading passage adding evidence about electricity, or creating another demand? It doesn't turn this blueprint into a regulatory requirement.
Ordinary classroom adjustments and formal access arrangements for regulated assessments or statutory tests have separate rules. This blueprint grants no formal arrangement and prescribes no application process. Check the responsible body's rules for the actual assessment and year. For whole-school timing, use the assessment calendar guide; for choosing evidence across phases, use the phase-specific planning cards.
Reopen the blueprint when evidence arrives
In the invented response record, a pupil names the components but leaves the conductor explanation unfinished. Ms Lewis keeps that gap. Attendance records show a missed lesson; the dated teaching log identifies it as the materials investigation. An assessment note records that the usual communication board wasn't available. These sources change the enquiry, but neither absence nor missing support proves why the response is unfinished.
At the 16 November review, Ms Lewis asks what the pupil had an opportunity to learn and communicate. She restores the missed teaching and agreed support, checks the pupil's account and plans another conductor task for 20 November. Mr Owen challenges whether the original task sampled observation-based reasoning at all. They retain version 1 and add version 2, recording the changed prompt, reason, reviewer and date. Outcomes from changed tasks can't be treated as equivalent measures of progress.
The family receives a limited account: what was demonstrated, what remains uncertain and what will be checked next. Leaders receive the coverage and unresolved teaching question, with missing evidence visible. Suppress or aggregate small-cohort patterns where identification is possible; a difference may be unstable and may not generalise. Keep sensitive context within authorised access and the school's retention arrangements.
No response here can diagnose, establish ability or effort, assign a fixed label, group pupils automatically, trigger a referral, calculate a risk score or justify a high-stakes recommendation. Later improvement cannot prove that teaching, support or software caused it. Curriculum exposure, delivery, practice effects, maturation, task differences, measurement error and missing evidence remain part of the review.
Keep the planning record beside the saved result
Where available to authorised staff, Student Radar's Exam & Assessment Windows supports scheduled sittings and candidate context. Digital Markbook records classroom marks, notes and missing reasons while keeping each assessment's scale visible. Use those records alongside the teacher-maintained blueprint and original evidence. They don't create or validate this blueprint, establish equivalent scores or decide the teaching response.
Before writing the next paper, take one taught sequence to a colleague and check one row together: content, demand, access, window and intended claim. Keep the unresolved question visible until its reviewer has the evidence. An assessment-focused walkthrough can show the product records and their limits.
Sources and further reading
Sources checked 10 November 2026. This is a classroom planning example for England.
- National curriculum in England: science programmes of study, Department for Education, published 11 September 2013; updated 6 May 2015. Year 4 electricity and pictorial representations.
- National curriculum test development handbook, Standards and Testing Agency, published 3 October 2024; updated 8 May 2025. The linked 2024 handbook describes national-test development, not a classroom mandate.
- Ofqual Handbook, Section D: General requirements for regulated qualifications, Ofqual. Handbook published 12 October 2017; manual updated 11 September 2026. Conditions D1 and D2 concern awarding organisations; the manual date doesn't imply each condition changed.
- How to ensure a test is valid, National Foundation for Educational Research, Assessment Hub. No publication or update date displayed; accessed 10 November 2026.