Exam Day Protocol Extended

Provenance: written 2026-09-29, against research/veridian/WMA12-verified-facts.md (read in full, §1–§9, specifically §3 for the rubric/cover-page wording and §5 for the archive's real scope — 20 series with question paper + mark scheme reviewed, June 2019–January 2026), content/veridian/technique-hub/wma12/forge-exam-technique-guide.md (read in full — this document builds directly on that one's Part 1 timing budget and Part 5's seven-point protocol, and does not restate either), content/veridian/technique-hub/wma12/signal-frequency-and-boundaries.md (read in full, for the one fully question-by-question-mapped real series it carries, June 2025), the 12 real lesson files in content/veridian/curriculum/wma12/ (grepped for existing timing/checkpoint content — paper-anatomy.ts already carries the standing 72-second-per-mark figure and is cited, not re-derived), and content/veridian/question-bank/wma12.ts (checked, no exam-day-protocol content present, no duplication risk). No new primary-source PDFs were opened for this pass — every claim below traces to a citation already verified in one of the sources above; where a search for a specific claim (reading time, a Pearson-recommended question order) came back empty, that absence is stated directly rather than filled in. A verification pass on 2026-09-29 re-checked every quoted fragment against the sources above and trimmed four quotes that had run past this course's own ~15-word limit on verbatim exam-board/mark-scheme text (the multiple-attempts rule, the January 2024 Q5(i) summation-button line, the June 2025 Q3 "purest route" line, and the June 2025 Q5(b) redone-trapezium line) down to short attributed fragments with the surrounding fact paraphrased — no factual claim changed.

#WMA12 — Exam-Day Protocol (Extended)

#What this document is, and isn't

forge-exam-technique-guide.md's own Part 5 already gives a seven-point exam-day protocol, and it is not wrong — it is deliberately narrow. This document is the deeper layer the task brief asked for: concrete reading-time use, a question-order strategy, real timing checkpoints built from actual per-question mark data, and a specific decision procedure for what to do if you're running behind. It does not re-derive the paper's basic shape (90 minutes, 75 marks, 72 seconds/mark average) — that is paper-anatomy.ts's standing reference, cited here by number, not repeated.

Two disciplines carried over from every other document in this technique hub: nothing below states a fabricated likelihood or grade boundary, and every piece of paper-specific advice is traceable to a real rubric line, mark-scheme mechanism, or examiner-report sentence. Where a piece of advice is sound general exam practice rather than something this paper's own sources establish, it is labelled [general practice] rather than left to look paper-specific.

---

#Part 1 — Before you start writing: what the cover page actually gives you

There is no confirmed separate reading time on this paper. This was checked directly against the facts bank's own rubric section (§3) and every cover-page quotation reproduced anywhere in this technique hub: no series reviewed carries a "you will have N minutes reading time before you may begin writing" instruction of the kind some other qualifications use. The confirmed cover-page instructions are: the paper's title and duration ("1 hour 30 minutes"), the total-marks line, the calculator-permission wording, and the single rubric sentence "Students must answer all questions." The 90 minutes is not split into a reading phase and a writing phase by Pearson — it is one block, and any "reading time" you take comes out of that same 90 minutes. Treat any exam-day plan that assumes a separate, protected reading period as wrong for this specific paper.

What the cover page does give you, confirmed (facts bank §3): the total mark tariff (75), the calculator-permission wording, and the instruction to have the Yellow formula booklet with you. It does not give you the question count in advance from the cover alone — that only becomes visible once you turn the page, and the confirmed real range is 9 to 11 questions (facts bank §3, independently re-confirmed in paper-anatomy.ts).

[General practice] — since there is no protected reading period, the practical equivalent is a fast first pass through the whole paper before committing to Question 1: turn every page once, note each question's own printed mark allocation, and only then start writing. This is not something a Pearson document instructs; it is offered here because two confirmed, paper-specific facts make it worth doing on this paper specifically, not generic advice repeated regardless of context:

1. Mark distribution per question is confirmed genuinely uneven and not predictable from position. The three real series mapped question-by-question in this technique hub (January 2021, January 2024, June 2025 — see Part 2's table below) show mark totals per question ranging from 3 to 14, and the heaviest question in January 2024 was Q9, not the closer or the opener.
2. Topic order is confirmed not fixed. forge-exam-technique-guide.md (Part 1) already established, from a direct read of the June 2025 examiner report, that Proof — which anchors Question 1 in most series reviewed (facts bank §6.1) — was Question 10, the paper's last question, in June 2025.

A candidate who has looked at every question's mark allocation once before starting knows, within the first two or three minutes, whether this sitting looks like a "January 2021 shape" or a "January 2024 shape" (Part 2) — and can budget accordingly instead of discovering a 14-mark question with fifteen minutes left.

---

#Part 2 — Question-order strategy

#The one thing that is confirmed, and the one thing that is not

Confirmed: "Students must answer all questions" (facts bank §3) — there is no optional or choice structure, so a question-order strategy on this paper is only ever about sequencing, never about selection. Every question must eventually be attempted; the only real decision is which order.

Not confirmed: no source reviewed for this course — not the spec, not any mark scheme, not any examiner report — contains a Pearson-published recommendation about attack order (e.g. "attempt questions in order" or "start with your strongest topic"). Any specific attack-order rule below is therefore [general practice], applied to this paper's own confirmed structure, not a Pearson instruction.

#What the confirmed structure actually supports

Three real series now have a complete, independently verified question-by-question mark breakdown in this technique hub:

All three put the single heaviest question in the second half of the paper — Q9 in both January 2024 and June 2025, Q10 in January 2021 — never in the first half, and January 2024's heaviest question is more than four and a half times its own lightest. [General practice], drawn directly from this confirmed shape: don't assume the front of the paper is the "easy half" and the back is the "hard half." Read each question's own printed marks as you reach it and adjust your remaining time budget against the marks still on the table, not against how many questions are left.

#Skipping a stuck question — general practice, applied to a confirmed mark mechanism

[General practice] — if a question genuinely stalls, move on and return to it, rather than staring at one line. This is generic test-taking advice, but it interacts with one confirmed WMA12-specific mechanism worth naming explicitly: the mark scheme's own multiple-attempt rule (facts bank §4) — an attempt left uncrossed-out is the one marked, and if every attempt is crossed out, or none are, the examiner marks all of them and takes "the highest single attempt." If you leave a partial attempt on a stuck question to come back to later, that partial working is still live and markable exactly as it stands if you never return to cross it out — leaving it visible costs nothing, and Part 3 of forge-exam-technique-guide.md already covers the mechanics of what counts as a scorable partial attempt (an M1 that doesn't depend on an earlier number, for instance) if you want to bank a partial method mark before moving on.

#Proof: worth naming as a genuine order exception

Proof (spec item 1) is confirmed as low-mark relative to most other topics on this paper — the signal-frequency table (signal-frequency-and-boundaries.md Part 1) records it at 2–4 marks per instance across six named series, well below the circle (7–10 marks) or differentiation (4–11 marks) — and it "usually opens or closes the paper" per that same table, having been confirmed at Q1 in June 2019, October 2022 and October 2023, but at Q10 in June 2025 and Q3 in June 2022. [General practice]: because it's genuinely low-mark wherever it sits, a Proof question is a reasonable one to attempt early regardless of its position in the running order, purely on a marks-per-minute basis — this is arithmetic on confirmed mark data, not a claim that Pearson orders papers this way.

---

#Part 3 — Timing checkpoints

#How this table was built, and what it is not

The three fully-mapped real series in Part 2's table make it possible to compute, honestly, what a candidate keeping exact pace through each of those specific real papers would look like at each question boundary, using the same 72-seconds-per-mark rate paper-anatomy.ts and forge-exam-technique-guide.md already establish as the paper's real average (90 minutes ÷ 75 marks). This is arithmetic applied to three confirmed real mark distributions — it is not a Pearson-published pacing table, and no fourth series has been checked to see whether these three are representative of the full 9-to-11-question range. Use it to see what "on pace" concretely looks like on three real papers, then rebuild the same sum for whatever paper is actually in front of you, from its own printed marks — the method is what's reusable, not these three specific rows.

Mapped onto the three real per-question sequences (cumulative marks after each question, then the pace-clock time that cumulative total implies):

How to read this table under real conditions. At 90 minutes you should have all 75 marks' worth of questions attempted on any of these three shapes — but the rate of arrival at that total is not linear. On January 2024, a candidate on pace is barely a fifth of the way through the paper's marks (21 of 75) after finishing Q4, because that series front-loads four light questions (3, 3, 6, 9); the same candidate should already be past the halfway mark on marks (46 of 75) by the end of Q7. On January 2021 and June 2025, the pace clock moves more evenly. The practical use of this table is not to memorise these three specific rows — it's to build the same cumulative-marks-vs-clock check for your own paper in the first few minutes (Part 1), so that at, say, the 45-minute mark, you know whether you should have roughly 35–40 marks' worth of questions behind you or noticeably more/less, given the actual shape of the paper you're sitting, rather than discovering the mismatch only at the end.

#What this table cannot tell you

It assumes uniform difficulty within a mark tariff, which is not true — a 6-mark question that is mostly a single method mark plus a long correct-technique chain (common in trapezium-rule or circle questions, per codex-formula-and-method-bank.md's method-skeleton entries) can take longer per mark than a 6-mark question built from several short, independent B marks. Treat the checkpoint as a warning system for a large drift (more than roughly 5–10 minutes off pace by the midpoint), not a precise per-question timer.

---

#Part 4 — What to do if you're running behind

Every item below is grounded in a real, confirmed cost — either a mark-scheme mechanism or an examiner-report sentence — not general "stay calm" advice, except where marked [general practice].

1. Do not switch to a calculator-only shortcut on a question that demands shown method, even under time pressure — it is confirmed to score zero, not partial credit. This is the single most counter-intuitive real finding in this technique hub for a candidate who is behind schedule and tempted to save time: forge-exam-technique-guide.md Part 2 cites two independent, real instances — January 2024 Q5(i), where the examiner report records candidates who reached for their calculator's summation button instead of the sum-to-infinity formula, and even when the right value came out, "no marks were scored"; and January 2025 Q5(a), where the mark scheme states outright, "Note that awrt 18.8 with no working is 0 marks." Running behind makes this trap more tempting, not less relevant — the time you think you're saving by skipping the method line buys you nothing if the mark scheme has nothing to read on the M line.
2. If a question offers more than one mathematically valid route, take the shorter one — this is confirmed to matter under real time pressure, not just in theory. June 2025 Q3's examiner report (research facts bank §6.9) verified three distinct valid strategies for the same area-between-a-curve-and-a-line question and explicitly flagged the algebraically "purest" one — subtract the two equations first, then integrate once — as "a more time-consuming approach" than the alternatives, needing several extra terms integrated or combined along the way. If you recognise more than one way into a question and you're behind, the mark scheme rewarding any valid method (facts bank §4) means the faster valid route is the correct choice, not a compromise.
3. Never recompute a part the question has already told you to reuse — this specifically costs both marks and time simultaneously, confirmed on a real paper. June 2025 Q5(b), verified directly: some candidates re-did the trapezium rule in part (b) despite the question insisting they use their answer to part (a), and the examiner report notes this scored nothing and was "also... detrimental to their timing." Recognise the specific phrasing that triggers this — "using your answer to (a)," "hence" — and go find that earlier number rather than re-deriving it, especially once you're already behind.
4. In the last few minutes, prioritise isolating and labelling answers you already have over starting new working — this is a confirmed real mark-loss pattern, not just intuition. forge-exam-technique-guide.md Part 4 records a real cost for the opposite habit: January 2024 Q2's final mark was commonly not awarded to candidates who "failed to isolate the required term from a list." A correct value sitting unlabelled among rough working when time runs out is a mark already earned and then lost for want of ten seconds circling it — a better use of a last few minutes than a fresh attempt at an unstarted question with no realistic chance of a complete, checkable method in the time left.
5. A partial, uncrossed-out attempt on a question you couldn't finish is still markable — leave it, don't erase it. Confirmed mechanism, facts bank §4: unless it is crossed out, an attempt is marked, and if you have two live (non-crossed-out) attempts the higher-scoring one is taken. Time spent tidying or erasing a genuine partial attempt in the closing minutes is very likely time spent for zero marginal benefit, and in the worst case removes a scorable M1 that was already on the page.
6. [General practice] — if you reach the final two or three minutes with one question completely unattempted, write down whatever you can even without full confidence: an equation set up but not solved, a diagram, a formula quoted (per facts bank §4's "quote the formula first" rule, quoting an appropriate formula can itself earn a method mark by implication even without a completed calculation). This is ordinary exam practice, not a WMA12-specific finding, but it interacts directly with a confirmed mechanism (M marks for "knowing a method and attempting to apply it," facts bank §4) — a bare, correctly-quoted formula is not nothing under this specific mark scheme, even with no numbers plugged in yet.

---

#Flagged gaps — what this document does not claim

• No confirmed Pearson-published reading-time allowance. Checked directly against the facts bank's rubric section and every cover-page quotation available in this technique hub; none exists. This document's Part 1 first-pass recommendation is offered as general practice specifically because no protected reading period exists, not as a description of one.
• No confirmed Pearson-published question-order recommendation. Every attack-order suggestion in Part 2 is labelled general practice applied to confirmed mark-distribution data; none is a paper-specific instruction from a spec, mark scheme, or examiner report.
• The timing-checkpoint table (Part 3) covers only three series out of the 20 in the archive, because only three have a complete, independently-verified question-by-question mark breakdown anywhere in this technique hub. Whether they are representative of the full confirmed 9-to-11-question range is not established — building the same table for a fourth or fifth fully-mapped series would be a natural next step, not something this pass could responsibly claim without doing it.
• Item 6 in Part 4 (a bare quoted formula scoring a method mark by implication) is a general application of the facts bank's own quoted rule, not tied to a specific named question where a candidate scored exactly this way with zero further working — flagged honestly as a reasonable inference from the rule's own wording, not a separately confirmed instance.
• This document does not cover wellness, sleep, or exam-morning logistics (what to eat, when to arrive) — this pass found no WMA12-specific grounding for any of that, consistent with forge-exam-technique-guide.md Part 5's own note that it found none either, and manufacturing generic wellness advice to fill space would violate this course's standing no-padding rule.