The IQA guide

Level 1 in particular

Three things about 7202-01 have no equivalent at Level 2, and one of them is the most checkable thing in either course.

The grade is the attempt

On a graded practical the three columns on the task sheet are headed by third attempt, by second attempt and on the first attempt. That is Pass, Merit and Distinction, and nothing else separates them — every criterion has to be met for any of the three.

So the attempt is not extra detail, it is the evidence for the grade. A Merit with no attempt recorded is a grade with nothing behind it, and it is the commonest gap in a Level 1 folder because nothing on the page asks for it twice. The cell asks for both, and says so when they disagree — "the sheet records Distinction on attempt 2, but attempt 2 is a Merit".

It warns; it does not block. You are sampling the assessor's decision, not making it. A Distinction recorded on the second attempt wants an action naming the discrepancy, not a module that refuses to record what is on the paper.

The unit grade is arithmetic, and it is checked

A unit is one assignment made of several tasks, and its grade is worked out rather than judged:

• Pass scores 1, Merit 2, Distinction 3.
• Total the points and divide by the number of graded tasks.
• Pass/fail-only tasks score nothing and are left out of the division — but still have to be passed.
• 1 to 1.5 is a Pass, 1.6 to 2.5 a Merit, 2.6 to 3 a Distinction.

This is the one place on a Level 1 folder where an assessor does sums, and the one place a wrong answer is invisible, because a number that has been worked out looks like a number that has been worked out.

Unit 109 is the case in point. Five tasks, three of them graded. A learner takes a Distinction, two Merits and two passes: 7 points over 3 graded tasks is 2.33, a Merit. Divide by five instead — the obvious mistake, and the one the sheet invites — and 1.4 is a Pass. A whole grade, gone, with nothing on the page to say so.

So every sampled cell carries its unit: each task in it, the grade and attempt recorded against each, the division written out, and a line saying how many pass/fail tasks were left out and why. Under it you record the grade on the assessor's mark sheet, and when the two disagree the module says which one the arithmetic supports.

It never states a unit grade off a partial sample. Nobody samples every task, so until every task in the unit has a grade recorded it reports what it has and stops — three Distinctions out of five tasks is not a Distinction, it is three Distinctions. And an open action holds the unit whatever the average comes to, because every task must be passed.

Three kinds of task, three thresholds

Practical	Observed. Every criterion must be met. Most are graded; a few are pass only.
Identification	100% — every item correct. Pass only.
Knowledge	Closed-book test under supervised conditions. 80% to pass. Pass only.

The cell says which kind it is and what the threshold is, because the three are marked differently and it is the sort of thing that gets remembered wrong.

Where the checklists come from

Every criterion on a Level 1 cell is the assessor's own marking sheet, task by task, taken from the C&G Candidate Task Manual — so "measured and joined pipes to +/- 4mm" appears here because it appears there. Each cell also carries the candidate's pre-assessment declaration and the sign-off block, because an unsigned declaration is the other thing that goes missing and nothing else looks for it.

What does not apply at Level 1 yet

The agreed figures feature under Standardisation covers both qualifications: fourteen measured tasks on 6035 and eleven on 7202. CP 1 says "to the dimensions given in the diagram provided", so the ±4 mm is stated by C&G and the sizes are yours — agree them at standardisation and the figure shows on the cell as you sample it, with the meeting and date it was agreed at.

1 · The risk profile

How the number is reached

Five factors, each scored 1–4 and weighted. Assessor experience, past performance and assessment judgement count double; cohort size and EQA standing count once. That is a score out of 28.

The rate is 10% plus 3.5% for every point above 8, capped at 80% and rounded to the nearest 5. A plan scoring 8 samples at the 10% minimum; one scoring 18 calculates 45%; one scoring 28 hits the cap.

Four rules can raise it, none can lower it

• New assessor, first time assessing — minimum 50%
• Significant past performance issues, still open — minimum 50%
• EQA action plan or sanctions in place — minimum 75%
• High-judgement assessment with a developing or new assessor — minimum 50%

The fourth rule is a Level 2 rule in practice: it needs high-judgement assessment, and 7202 is mixed. A Level 1 plan with a new assessor is still floored at 50% by the first rule.

So a plan can score 18, calculate 45%, and show 50% — the fourth rule caught it. The rationale names the rule that applied. Why: a number an EQA cannot trace is a number they will ask about, and "the system said so" is not an answer.

The same team, two different rates

Assessment judgement is a fact about the qualification, so it is the factor that separates the two courses. One assessor rated experienced with no issues, one a medium cohort, EQA in good standing:

Level 1	judgement mixed — score 13, rate 30%
Level 2	judgement high — score 15, rate 35%

Nothing else changed. The factor counts double, so it moves the score by two and the rate by seven per cent before rounding.

Three of the five answer themselves

Assessment judgement is stated by the qualification — high for 6035, mixed for 7202. Assessor experience and past performance are derived from the ratings you give each assessor individually, taking the most cautious amber-or-below rating.

Why not an average: averaging under-samples the weaker assessor and over-samples the stronger one at the same time. Why not the worst outright: that takes the whole team to 100% the moment one person is red, which turns a remedy for one individual into the blanket sampling §3.2 warns against.

A red-rated assessor raises its own targeted instruction instead — verify every decision they made — which is proportionate, evidence-based, and temporary.

Rating an assessor

Two answers each: experience (expert → new) and past performance (no issues → significant and ongoing). Those give a RAG. Not rated is a fourth state and looks unfinished on purpose — somebody nobody has rated is not somebody known to be low risk.

If anyone is unrated the plan will not derive at all and falls back to its own saved answer, saying so. Why: sizing off the rated subset would quietly under-size, and under-sampling is the one direction not to err in.

When the number is wrong

There is a manual override. It asks for a reason and the reason goes in the audit pack. It exists for the case where an EQA or centre policy requires a specific rate — not as a way of disagreeing with the arithmetic. If you find yourself using it routinely, the factor answers are wrong rather than the formula.

What the module does on your behalf

The rationale includes an adaptive strategy section — four claims the module makes about itself, written so you can defend them:

• Learners over an error-rate threshold are flagged for increased sampling.
• An assessor observation triggers when a share of the cohort is flagged, at a lower threshold for a developing assessor.
• Resampling of adjacent units is prompted when an action is resolved, to confirm corrected practice held.
• The phase — pre-formative through post-summative — changes whether outstanding actions are advisory or urgent.

Why it is written down: an EQA asking "how do you respond when practice slips?" wants a described process, not an assurance.

2 · The CAMERA record

§2.2 requires documented evidence that the sample was planned on the CAMERA principles, and §3.7 that the reasoning is recorded in the rationale. This is that evidence, and it keeps itself.

C	Candidates — learners across three visits and every unit group
A	Assessors — rated, qualification-checked, and every sample records who assessed it
M	Methods — at Level 2, direct observation and product evidence. At Level 1, practical observation, identification and knowledge tests. All of them sampled at each visit.
E	Evidence types — a portfolio at Level 2; externally set assignments, marked internally, at Level 1
R	Records — decisions, feedback, staff credentials, standardisation, learner compliance
A	Assessment — internally assessed at Level 2; externally set and internally marked at Level 1. Sampled at interim and summative either way.

The M and the E come from the qualification, so the record says whichever is true of the plan you are on. Each letter reads covered or partly covered, and a partly-covered one says exactly what is missing and links to it. The header counts how many of the six are there.

A missed event takes a letter back down. A visit date that passed with nothing sampled, an observation date nobody kept, a standardisation meeting with no minutes — each of those reverts the letter it belongs to. Why: otherwise the record describes a plan rather than what happened, which is the difference between a document and a claim.

Nothing in the future counts against you. A visit booked for June is not overdue in September.

3 · The sampling grid

What a column is

At Level 2 a column is one of the 41 portfolio sections. At Level 1 it is one of the 31 assignment tasks, grouped under its unit. Why a task and not a unit at Level 1: ten columns draws no diagonal worth having across a class of twelve, and a unit is only achieved when every task in it passes — so the task is the thing being sampled.

How the plan is drawn

One cell in every 100 ÷ rate, laid diagonally across learners and units. At 50% that is one in two; at 25%, one in four. Why diagonally: it covers every unit across the cohort rather than the same units on everybody, which is what makes the sample representative instead of convenient.

Three visits

The diagonal is cut into three roughly equal parts. At Level 2 that is safety and processes first, then everything up to the cylinder, then the rest — 14 / 14 / 13. At Level 1 it follows the units in delivery order, 101 through 113 — 11 / 11 / 9. Each visit takes a different part of the folder, so each samples different skills and knowledge — and because learners work at different speeds, each visit catches them at a different point in their programme. That is §3.1's range of stages.

You can pin a unit group to a particular visit; everything else rebalances around it. Sampling earlier or later than planned is expected, not an error. The plan records where the work actually went.

Opening a cell

Each cell carries the checklist for that unit, page by page, as the assessor's own paperwork lays it out. You also record:

• Who assessed it. Required on any green or orange decision where the plan has more than one assessor. Why: a judgement that does not say whose work it was cannot evidence that every assessor was covered.
• The outcome — planned, satisfactory, action required, or pending.
• Your initials and the date.
• A variation reason if the cell was not in the original plan.
• At Level 1 only: the grade on the sheet and the attempt it was reached on, and the unit's standing so far. See Level 1 in particular.

Feedback written for you

Mark anything on the checklist as N and the feedback field fills itself with the specific failures, page and item by page and item — "P8 pg 41: Sizes correct — N". Why: feedback that names the page is feedback an assessor can act on, and it is the difference between a record that improves practice and one that logs a complaint.

Agreed figures on the measured tasks — Level 2 for now

The task manual sets the tolerance — measurements within ±3 mm, angles within 3 degrees — but the sizes come from your own pipe detail or task drawing. So an IQA could see the assessor ticked "within 3 mm" with no way of knowing 3 mm of what.

Fourteen 6035 tasks now carry the figures agreed at standardisation, shown on the cell as you sample it. Where nothing has been agreed the cell shows what the sampling checklist already holds, marked as a suggestion rather than a decision.

Eleven 7202 tasks carry the same thing — CP 1 says "to the dimensions given in the diagram provided", so the ±4 mm is stated by C&G and the sizes are yours. The tasks are spread across the first three meetings in delivery order, so no single agenda has to agree eleven sets of dimensions at once.

Countersigning

The sign-off asks for a countersignature when the assessor who made the decision is working towards their award, is rated developing or new, or is not cleared for that qualification. It follows the person recorded on the cell — so on a team of an expert and a first-timer it asks on the right work.

Evidence sent back

An assessor can attach evidence against an action; the cell then sits as pending until you approve it or return it with feedback.

5 · Who may assess

Every assessor and IQA gets a verdict per qualification, because the three handbooks ask for different things.

7202-01	TAQA is "valued … but not currently a requirement". It also says centre staff may hold more than one role but cannot internally verify their own assessments.
6035-02	Assessors must hold, or be working towards, TAQA
8202-35	No assessor award named; asks for recent relevant teaching and assessment experience

The matrix answers per qualification, so the same person can read differently on each. Somebody working towards their award is clear for Level 1, where the award is not required, and countersign for Level 2, where it is. That is the matrix doing its job, not a contradiction.

The trade qualification carries a level

Competence must be at or above the level assessed — and what a certificate evidences is not simply its own level. An NVQ is the competence qualification: a Level 2 NVQ reaches Level 2. A Diploma is a study qualification — 6035-02 says in terms that it does not make somebody a fully qualified plumber — so a Level 2 Diploma reaches Level 1, and a Level 3 Diploma reaches Level 2.

Legacy Craft and Advanced Craft count only with a CPD record beside them, which the handbook requires.

The assessor award splits by setting

ACWE is workplace only. AVRA is the learning environment. CAVA is both, and the legacy A1 matches CAVA most closely. All three of these courses are centre-based, so ACWE alone does not reach them and reads wrong award.

The four verdicts

• Clear — qualified and occupationally competent for that level.
• Countersign — working towards the award, which is accepted; the decisions need a second signature.
• Below level / Wrong award / Not cleared — and the sentence says which and why.
• No records — nothing on file. A gap in the record, not a judgement about the person.

Competence, and the four routes to it

6035-02 lists four and says "one, preferably more". The one an EQA expects is the trade qualification; the others read clear, with a caveat because you will be asked to defend them:

• A CPD record together with a current industry registration — the handbook joins them with an "and", and both must be in date.
• A verified CV — it counts once somebody is named as having checked it, with a date. Of the four routes it is the only one the person writes about themselves.
• Standardisation attendance, which evidences knowledge of the qualification standards.

A trade qualification below the level is not rescued by a fallback route. The centre has stated that person's occupational level and it does not reach.

An uncleared assessor who has already assessed sampled units is named, with the count. Why: that is the first thing an EQA reaches for.

6 · QA records

Dates

Three sampling visits, four standardisation meetings and an assessor observation. Set them and reminders go to both the IQA and the assessor seven days before and on the morning.

One meeting before the year, three inside it. The first agrees what good looks like on the units being delivered first — which only works if it happens before anybody assesses against it. The second, at the end of term one, checks the standard held and sets the next one for the units coming up. The third does the same for that second set, with two thirds of the year sampled and time left to act on what it found. The fourth closes the year off before the EQA arrives. Why four rather than three: a standard agreed in September and not revisited until March is an assumption, and the units being assessed in March were never the ones the September meeting looked at.

Each meeting opens with the items that belong to it rather than the same two groups every time, and the agenda shows its own running length as you build it — nine items, about ninety minutes — so an over-long meeting is visible while you are still choosing rather than at four o'clock on the day.

Standardisation

Each meeting has an agenda and a record. Fixed items — opening, last minutes, matters arising — cannot be removed, because they are the procedural frame an audit expects. Standard items are pre-selected and optional ones can be added. Attendees sign with initials.

Two standing items carry the agreed figures for the measured tasks on 6035: unit 204's nine processes tasks at meeting 1, the installation tasks at meeting 2. Why split: the processes figures have to be settled before anybody assesses against them, and one agenda carrying all sixteen is one people skip.

Agendas and records both export to PDF.

Staff documents

CV, assessor or IQA award, trade qualification, CPD, industry registration and teaching qualification, per person. Anything can be marked held offline — the module is not your filing cabinet and will not argue with you about a certificate that is in one. CPD and registrations carry a renewal date and go amber, then red, as they age.

You can add your own rows. The two that decide a verdict cannot be removed, because a matrix editable until it says everybody is fine is not a control.

Observation of assessment practice

A structured record in four parts — pre-assessment check, the assessment process, post-assessment and feedback, then your summary and action plan — graded, with actions and a follow-up date. Exports to PDF.

Why it matters to CAMERA: the matrix evidences that somebody is qualified. The observation evidences that they are currently competent. A certificate from 2009 says what somebody was; watching them assess says what they are.

7 · Action lists

Everything outstanding, split by who owes it — you or the assessor. Actions arrive from sampled cells marked action required, from CAR3s, and from pattern findings. Resolving one prompts a resample of adjacent units, to confirm the corrected practice held rather than the single cell being fixed.

8 · Reports and the EQA

The audit pack

One PDF: the risk assessment and its working, the CAMERA record, who may assess, key dates, the sampling grid, compliance checks, the assessor observation and your notes for the EQA. Why it carries CAMERA and the staffing matrix: §2.2 asks for documented evidence, and a screen the EQA never sees is not a document.

Folder sheets

One per learner — sampled progress, mandatory checks and sign-off, your comment and any actions. This is the sheet that goes in the front of the portfolio.

Assessor coverage

One line per assessor: how many of the samples taken are of their assessing, and the sentence that matters — "you have not yet looked at anything B assessed." There is deliberately no coverage score, because a target is what people chase instead of the question it stands in for.

A plan whose samples record no assessor at all reads as a recording gap, not a coverage gap. The two have different remedies and only one of them is an accusation.

Strategy document

A written statement of the sampling strategy, for the quality file.

Set up once

EQA access

Add the EQA's email and they get a live view of every group's plan, sampling progress, risk levels and observation outcomes — and you can record their feedback and any CAR3 against the plan.

Copying a record to other groups

A standardisation record that covers several groups can be copied across rather than retyped.

How long records are kept

Under Reports → Retention, set how long this college keeps its IQA records — three, five or six years, running from the end of the academic year a plan belongs to. Six is the default, because it is the longest period a college delivering funded apprenticeships is typically under.

Nothing is deleted automatically. A plan past the period is listed with what it holds — how many learners are named, how many decisions recorded — and you export it or delete it yourself. Deletion asks you to type the group name back, because it cannot be undone. Why it works that way: the college is the data controller and PlumbMate is the processor, so how long your records are kept is your decision, not ours. The retention policy sets out the obligations behind the choice.

Things that handle themselves

No action needed — listed so you know what you are looking at.

• Two people editing at once. The plan shows who else is in it and keeps both sets of changes.
• A transferred learner is flagged on their cells, because their samples may be out of sequence.
• Withdrawn learners keep their cells. A sample that was taken is evidence. They stop counting towards this plan's figures.
• Export grid does what it says.

📚 Back to the tutor guide →