Standards verification: what Pearson samples and how to be ready
Standards verification (SV) is Pearson checking your centre's internal standard: an appointed Standards Verifier reviews a sample of your assessed learner work and — crucially — the records behind the decisions. Centres that prepare in September experience SV as a formality. Centres that prepare in May experience it as a crisis.
The mechanics
- Allocation. Pearson assigns a Standards Verifier to your centre for the cycle, per programme area.
- Sample request. The SV requests a sample of assessed work — learner evidence plus the assessment records and internal verification logs behind it, by a stated deadline.
- Review. They compare your decisions and records against the national standard for each unit sampled.
- Outcome. Either your standard is confirmed, or you get findings — required actions, possible re-assessment, and an effect on your centre's risk rating for future cycles.
What the sample must contain (per learner, per unit)
- The assignment brief actually used (correct version — see the PSA guide)
- The learner's evidence, complete and as-submitted
- The assessment record: per-criterion decisions with evidence locations
- The IV trail for that work: what was sampled, agreed, signed off
- Any resubmission paperwork: authorisation, first attempt, final outcome
Missing pieces are findings in themselves — "good teaching, incomplete records" is the most common SV report line in existence.
The five findings that repeat every cycle
- Assessment records that summarise instead of decide. "Strong project, Merit" verifies nothing. Per-criterion verdicts with locations do.
- Briefs that don't match the spec. Criteria changed, added or omitted by a well-meaning author — stage-1 IV exists to stop this.
- Resubmission rule breaches. Second attempts nobody authorised, or windows quietly exceeded — the rules are short, and the breaches are well documented.
- Verb-level leniency. Described work recorded as met against analyse criteria. The SV checks the verb against the evidence, every sample, every time — see command verbs.
- Unauthenticated work. Especially in the AI era, evidence with no authentication trail invites deep sampling — see the detection guide.
A preparation calendar that works
| When | Do this |
|---|---|
| September | Assessment plan per unit: briefs, versions, deadlines, named assessor + IV |
| Each unit, before IV | Run evidence through criterion-level checks; resolve gaps while learners are still in the room |
| After each round | Complete records at marking time; IV log signed within days, not months |
| March | Internal mini-audit: pull 3 scripts per unit and check the five findings against them yourself |
| Sample request arrives | Collation is a formality because every file already has its record attached |
The honest shortcut
Every finding above is about records, verbs and rules — not about teaching quality. A tool that produces per-criterion verdicts with cited evidence for every script (checkb.tech) makes the record as you mark, and its export is the paperwork the sample asks for: see the export format.
FAQ
What happens if the SV disagrees with our standard?
You'll get required actions — commonly re-assessment of sampled work and a report condition. There's an appeals route via Pearson's enquiries process; centres rarely need it when records are complete.
Does every centre get sampled every year?
Sampling is risk-based, but new centres and centres with prior findings should expect coverage every cycle. Prepare as if sampled regardless.