Technique Masterclasses

Statistics 1

Examiner Mindset Guide

WST01 (Statistics 1) — Examiner Mindset Guide

Provenance: written 2026-09-29, built entirely on research/veridian/WST01-verified-facts.md (compiled 2026-09-12, spot-check addendum 2026-09-25), specifically its §4 (mark-scheme conventions, transcribed verbatim from the "General Marking Guidance" and "Special notes for marking Statistics exams (for AAs only)" boilerplate) and §5 (per-topic examiner-report quotes). No new PDFs, web fetches, or external searches were performed for this document — everything below traces to that one facts file, cited by series and question code. Cross-checked directly against forge-exam-technique-guide.md §2 (which already transcribes the full M/A/B glossary, the general marking principles, the misread penalty and the AAs-only note in full) and forge-level-threshold- guide.md (which already states the mark-scheme-type finding below) and signal-examiner- intelligence.md (which already analyses how examiner-report prose is phrased) so this document does not re-run either of those. What follows is deliberately narrow: it takes the same primary-source sentences those documents already quote and asks a different question of them — not "what does the mark scheme say" and not "how does the report talk," but "in what order does a marker actually apply this, and what is the marker checking first." Where that narrower question turns out to have very little unmined material left in the record, this document says so rather than padding around it — see §4.

1. The marking process, confirmed rather than assumed: discrete M/A/B, not holistic best-fit

This is checked, not guessed. The "General Instructions for Marking" boilerplate opening every one of the 14 WST01 mark schemes reviewed defines three independent mark types — M (method), A (dependent accuracy, "M0 A1 is impossible"), B (independent, standalone) — plus follow-through (ft) marks, and states outright that "mark schemes should be applied positively" with "no ceiling on achievement." There is no level-descriptor language, no 1–5 band, and no instruction anywhere in the 14 reviewed schemes to award a "best-fit" holistic judgement of a whole answer the way a levels-marked, prose- based subject would. forge-level-threshold-guide.md independently confirms the same reading: "WST01 is marked M/A/B, not by level descriptors (1–5 bands, best-fit judgements) the way an essay-marked subject is." Every mark on a WST01 script is a discrete point on a numbered list, awarded or withheld against a specific, named criterion — never a single overall impression of the answer.

This matters for how a student should actually write an answer: a WST01 script is not read once for a general impression and scored against a rubric band. It is read line by line against a checklist, and each mark point is either present on the page or it isn't.

2. What the marker checks first: the decision order behind the "AAs only" note

Every mark scheme carries a short block headed "Special notes for marking Statistics exams (for AAs only)" — the mark schemes never define what "AAs" stands for, and this document does not guess at an expansion it cannot verify. What that block actually instructs, read as a sequence rather than as the flat list forge-exam-technique-guide.md §2.4 already reproduces, is a three-step decision order:

  1. Is there a correct method on the page, in any form? "Any correct method should gain credit" — the marker's first question is not "does this match the model solution's route" but "is this a valid route," full stop. A method that reaches the right relationship by a different path from the one printed on the mark scheme is still creditable.
  2. If the marker genuinely can't tell, they escalate rather than decide alone. "...please send to review" — the instruction is explicit that uncertainty is resolved upward, not by the individual marker's best guess. A marker is not trained to penalise an unfamiliar-looking method; they are trained to flag it.
  3. Once a method is accepted, the tolerance for a slip depends on which mark type it sits in. "...we generally allow or condone a slip or transcription error" in a method line, but "we do not, however, condone or allow these errors in accuracy marks." The same numerical slip is judged differently depending on whether it appears inside the working (forgiving) or in the number the question actually asked for (strict).

Read in that order, the note describes a marker who is trained to look for validity of process before anything else, to treat their own uncertainty as a signal to escalate rather than a coin to flip, and to apply a harder standard to the final figure than to the reasoning that produced it. That ordering — process first, escalation over guessing, and a harder line at the end than in the middle — is the "examiner mindset" this document was asked to describe; the individual sentences it's built from are already on the page in forge-exam-technique-guide.md §2.4, so this section's contribution is the sequencing argument, not new quotations.

3. Marking in practice: three real, cited instances of the mechanism firing

A live invocation of the misread rule. The general marking guidance states that misreading a given number costs "deduct two from any A or B marks," rather than zeroing the affected working outright. This is not a theoretical clause: on the real October 2021 series, the examiner report records that "an alternative mark scheme, applying the usual misread penalty, was used" after a candidate misread a given figure in Q3 — i.e. a marker actually built and applied a second, adjusted scheme for that candidate's script rather than either ignoring the misread or failing the whole question. That is the mark scheme functioning as a live decision procedure on an individual paper, not just as printed boilerplate.

A diagnosis that tracks one exact sub-step, not the answer as a whole. On January 2023 Q2(d), a conditional-probability question, the examiner report identifies the common wrong answers as two specific malformed expressions — both of which, checked directly against the real mark scheme, use the correct conditioning denominator (61/234, carried forward correctly from the previous part) but smuggle a spurious extra copy of that same fraction into the numerator. The point for this document is not the maths (that belongs to the topic lessons) but what it shows about how the script was read: the marker's own diagnosis locates the error at one specific sub-expression inside a multi-line calculation, not as a verdict on the final answer. That is only possible under discrete, line-by-line marking — a holistic best-fit reading of the whole response would have no mechanism for pinpointing which fraction, inside which line, was the spurious one.

A general comment naming what the marking is designed to separate. On June 2022 Q6(c), the examiner report describes the question as "the final part of the paper... also the most discriminating part" — a marker's-eye observation about which specific mark point on the paper is doing the work of telling strong candidates from average ones, made possible only because the mark scheme assigns that discrimination to an identifiable point (in this case, a conditional-probability step inside a Normal- distribution question) rather than to an overall impression of the script.

4. What this document does not do, and why

This facts bank has already had a genuinely thorough first pass taken through it by other documents in this same folder: the full M/A/B glossary and abbreviations table, the general marking principles verbatim, the misread penalty, and the AAs-only note in full already live in forge-exam-technique-guide.md §2; the mark-scheme-type finding in §1 above is independently stated already in forge-level-threshold-guide.md; and the question of how examiner-report prose itself tends to be phrased (register, headline general-comment placement, vague quantifiers, repeated findings across years) is the entire subject of signal-examiner-intelligence.md. This document does not re-derive any of those, and a reader wanting the complete verbatim mark-scheme text should go to forge-exam-technique-guide.md §2 rather than expect it repeated here.

The honest limit on anything new this document could add is the underlying research base itself: only 6 of the 14 WST01 mark-scheme series have a confirmed, independently-read examiner report (Jun 2019, Jan 2021, Oct 2021, Jun 2022, Jan 2023, Jun 2024 — WST01-verified-facts.md §3), and that base has already been mined for quotable material across roughly forty documents in this folder before this one was written. No grade-boundary data exists anywhere in the source record either (signal-frequency-and-boundaries.md §2 checked directly and found none), so, in line with this course's standing rule against inventing what wasn't checked, this document makes no claim about how marking-process rigour maps onto any grade boundary, and no claim about how often any single AA actually escalates a script to review — the source states the instruction to escalate, not a rate.

Sources

research/veridian/WST01-verified-facts.md §3 (source archive and the 6-of-14 examiner-report denominator), §4 (mark-scheme conventions, including the "General Marking Guidance" and "Special notes for marking Statistics exams (for AAs only)" boilerplate, transcribed there verbatim from Pearson's own mark-scheme PDFs), and §5 (Jan 2023 Q2(d), Jun 2022 Q6(c)). Cross-checked against content/veridian/technique-hub/wst01/forge-exam-technique-guide.md §§2, 2.4, forge-level-threshold- guide.md, signal-frequency-and-boundaries.md and signal-examiner-intelligence.md in this same folder, specifically to confirm this document does not repeat their content. All direct quotations above are Pearson's own wording as transcribed in the facts file, reproduced here only as short extracts under Pearson's own copyright, not in bulk.