Exam Day Protocol Extended

Provenance: written 2026-09-29, grounded in research/veridian/WMA14-verified-facts.md (11 series with at least one primary document reviewed; 4 full Examiners' Reports — January 2021, October 2021, October 2022, January 2024 — plus the January 2026 mark scheme and question paper, which has no examiner's report yet). Cross-checked against forge-exam-technique-guide.md (§1's real per-question mark-tariff table and §5's existing short exam-day list) and signal-frequency-and-boundaries.md (§1's topic-frequency table) so nothing below repeats either file — this document goes deeper on four things neither of those covers: reading-time use, question-order strategy on a compulsory-all paper, derived timing checkpoints, and a "running behind" playbook. No new primary Pearson document was fetched for this pass; every fact below traces to a citation already established in the verified-facts file, re-read and re-applied here specifically to exam-day pacing.

#WMA14 · FORGE — Exam-Day Protocol (Extended)

forge-exam-technique-guide.md §5 already gives a short, ten-line exam-day list. This document does not repeat it. It goes one layer deeper on the four things that guide's own §1 and §6 flag as needing more than a summary: how to actually use the time before you start writing, how to sequence a paper where every question is compulsory, what a real timing checkpoint looks like against this paper's own fixed totals, and — the genuinely new material — what the verified record says to do when you are behind schedule mid-paper.

Every specific claim below is either a direct citation (Month Year, Qn) traceable to WMA14-verified-facts.md, or explicitly marked as this document's own arithmetic or as general exam practice rather than a Pearson-sourced instruction. Where the real record has nothing to say, that gap is stated rather than filled.

---

#1. What is actually fixed before you open the paper

Confirmed directly from spec P4.2.2 and from real question-paper cover pages (§2 of the facts bank):

• 75 marks, 90 minutes, every question compulsory. No section choice, no optional questions.
• Question count is not fixed. Three real papers checked directly: June 2022 had 9 questions, October 2022 had 11, January 2026 had 9. Only the 75-mark total and the topic list are guaranteed — not how many questions it takes to reach 75.
• Required on the desk: the Mathematical Formulae and Statistical Tables (Yellow) booklet, and an ordinary scientific calculator. Appendix 6 bars any calculator with symbolic algebra, differentiation or integration built in, or with retrievable stored formulae — confirmed on the June 2022 cover page itself, not just the spec.
• The general rubric on unsupported answers: mark schemes state plainly that an answer with no working "may not gain full credit," with the general policy that only a step genuinely doable "in your head" is exempt.
• A single question can override the paper's general calculator policy. WMA14 is calculator-allowed throughout, but Jan 2026 Q7 prints, twice, that "solutions relying on calculator technology are not acceptable" — covering all four of its parts. Check every question's own printed instructions, not just the front cover, before reaching for a calculator shortcut.

---

#2. Reading time — what is real and what is general practice

There is no separately timed reading period on this paper. This is worth stating explicitly because some exams (certain English literature and comprehension papers, for instance) do allocate a distinct reading-only period before writing may begin. Nothing in the spec, the cover pages, or any mark scheme or examiner's report reviewed for WMA14 describes any such period here — the entire 90 minutes is exam time, and a candidate may begin writing the moment the exam starts. Treat "reading time" for this paper as a self-imposed budget carved out of the 90 minutes, not an officially allocated extra.

What is genuinely paper-specific, not general advice: because question count varies (9–11, confirmed above) and per-question tariffs vary far more than a flat average suggests — real confirmed items range from a 4-mark proof to a 16-mark combined parametric/normal/area question (Jan 2026 Q7, 2+5+6+3 across four parts) — a quick skim of every question's printed tariff before committing ink is directly useful on this specific paper in a way it would not be on a paper with fixed, identical question shapes series to series.

What is general exam-technique practice, not a Pearson-sourced instruction, flagged as such: spending roughly the first two minutes of the 90 on that skim — counting the real questions, reading every printed tariff and any question-specific rubric override (Section 1 above) — is standard test-taking advice applied to this paper's own confirmed variability, not a quoted examiner recommendation. No examiner's report reviewed tells candidates how to spend their first few minutes.

---

#3. Question-order strategy on a compulsory-all paper

Because every question must be answered, "question order" here does not mean choosing what to attempt — it means deciding, deliberately rather than by default, whether to work the paper in its printed order or to sequence around a question you can see is disproportionately large.

What the real record supports: the confirmed per-question sizes already tabulated in forge-exam-technique-guide.md §1 show the smallest real items at 4–6 marks (proof, binomial) and the largest confirmed real item at 16 marks across four parts (Jan 2026 Q7) — nearly three times the size of the average question implied by a flat 75-mark/9-question paper. A candidate who reaches a large multi-part question mid-paper without having budgeted for it is the concrete, paper-specific risk this variability creates: spending disproportionate time on one part of a 16-mark question can crowd out an entire smaller question later in the paper that would have taken a fraction of the time.

What this document adds, as an explicit inference rather than a quoted instruction: once the opening skim (Section 2) has identified which question is the largest and how many distinct parts it has, a candidate under time pressure has a genuine choice — work it in place, or bank the smaller, self-contained questions around it first and return to the large one with whatever time remains. Both are legitimate; the point is that this paper's own confirmed size variation (Section 1) makes that choice worth making consciously rather than defaulting to printed order without having looked ahead.

What is general test-taking practice, not paper-specific, flagged as such: "do the questions you find easiest first" is common advice for any compulsory-all paper and is not something any WMA14 mark scheme or examiner's report states or recommends. Nothing in the verified record says Pearson expects, rewards, or penalises any particular order of attempt — the compulsory-all rubric is silent on sequencing.

---

#4. Timing checkpoints — a derived tool, not an official schedule

forge-exam-technique-guide.md §1 already establishes that the flat 75-marks/90-minutes average (1.2 minutes per mark) is "the wrong pacing tool" on its own, because real tariffs vary from 4 to 16 marks. The checkpoints below are this document's own arithmetic against the paper's two fixed, real totals — not an examiner-published pacing schedule, and not a claim that Pearson endorses any specific checkpoint.

Why this table is a floor, not a plan. It assumes every mark takes the same time, which Section 1 and forge-exam-technique-guide.md §1 both establish is false for this paper specifically — a single confirmed real question (Jan 2026 Q7) is worth 16 marks and chains three separate techniques across four parts, which at the flat rate alone would already consume nearly a fifth of the entire paper's time budget. Use the checkpoints as a way to notice you are badly behind, not as a target to hit exactly question by question.

---

#5. What the record says to do if you are running behind

This is the genuinely new material in this document: concrete, cited responses to being short on time mid-paper, drawn from real mark-scheme mechanics and real examiner-report findings that are directly relevant to a time-pressured decision, even though none of the sources below frame themselves as exam-day advice.

• Quote the formula first, even on a question you won't finish. The mark scheme's own stated mechanism — "the formula should be quoted first" — means a method mark can be banked from that line alone, independent of what happens afterward. Under time pressure, writing the named formula for whatever technique a question opens with (the binomial series, the scalar product, $\pi\int y^2\,dx$, integration by parts, the cosine rule) is a real, cheap way to secure a mark even if you run out of time before finishing.
• Write the dependent stages you do reach, even out of order or incomplete. This paper's own schemes routinely split a derivation into two, three or four dependent marks (M1 → A1 → dM1 → A1* on the richest proof anchor; M1 → M1 → A1 → dM1 on the foot-of-perpendicular anchor) rather than one lump mark. A rushed answer that shows the intermediate stage genuinely reached is still gaining the marks attached to that stage; jumping straight to a guessed final answer forfeits marks that were otherwise reachable in the time you actually had.
• Don't let a struggled part (a) cost you an attemptable part (b). Oct 2021 Q7's examiner commentary notes that a substantial number of candidates did not attempt part (b) at all after struggling with part (a) — even though (b) is a separate, smaller 2-mark item building on the same position-vector idea rather than requiring a correct (a) to approach. If you're behind and a question has multiple parts, attempt every part rather than abandoning the rest because the first one didn't go well.
• Recognise a method that isn't converging, and change it, rather than grinding on. Oct 2022 Q5's report records that candidates who missed the efficient reverse-chain-rule route on that question "wasted much time pursuing incorrect methods" including integration by parts and a wrong log form. This is the one documented case in the reviewed record of time itself being named as the cost of a wrong method choice — the general lesson, stated here as this document's own inference from that one case rather than a repeated finding: if a method has run several lines without visibly approaching a usable form, that is a real signal to stop and reconsider the approach, not just push the same method further.
• A misread doesn't necessarily cost the whole question — finish it anyway. The real Oct 2024 Q1 scheme caps a specific misread (reading the binomial index's sign wrong) at B1M1A0A0 rather than zero. If you spot a slip partway through and are short on time, finishing the method correctly from the wrong premise is still worth marks — it is not a reason to abandon the question.
• Don't attempt proof-by-contradiction questions blank, even rushed. The general commentary records this as the topic most often skipped outright, and the setup step alone — stating a clear assumption — routinely banks a method mark before any of the harder logical scaffolding is needed. A short, incomplete attempt under time pressure is worth more than a blank answer.
• If you rewrite an answer and aren't sure which version is better, don't cross either out. The general marking guidance: when all attempts are crossed out, or none are, the examiner marks every attempt and scores the best one; only crossing out all-but-one restricts scoring to what's left visible. A hurried second attempt written in the margin, left uncrossed alongside the first, is marked as whichever one is stronger — crossing out the wrong one under time pressure is the actual risk.
• Keep exact values even when rushing. Mark schemes note that marks are normally lost for resorting to rounded decimals where exact or surd working was required — a habit that costs marks regardless of whether the underlying method was rushed or careful, so it is not a place to cut corners for speed.

---

#6. Flagged gaps

Stated plainly rather than papered over, matching this paper's own facts bank discipline:

• No separately timed reading period is verified for this paper. Section 2 above states this directly; nothing in the spec, cover pages, or any mark scheme or examiner's report reviewed describes one.
• No examiner's report reviewed explicitly names running out of time as the cause of a blank or omitted answer. Every "did not attempt" or "omitted" finding cited in the facts bank (proof by contradiction especially) is framed by Pearson's own examiners as topic avoidance or unfamiliarity, not as a stated time-pressure effect. Section 5's "running behind" advice is this document's own application of real, time-relevant mark-scheme and error findings to a time-pressured scenario — it is not itself a Pearson statement about time management.
• Section 4's checkpoint table is this document's own arithmetic against the paper's two fixed real totals (75 marks, 90 minutes), not an official Pearson pacing schedule, and it is explicitly a floor rather than a plan for the reason stated there.
• Section 3's "bank the small questions first" option is general test-taking practice, not sourced from any WMA14-specific rubric or report; only the underlying tariff-variability evidence it responds to is paper-specific.
• Evidence-base scope is unchanged from WMA14-verified-facts.md and forge-exam-technique-guide.md: 11 series with at least one document, 4 full Examiners' Reports (Jan 2021, Oct 2021, Oct 2022, Jan 2024), and the January 2026 mark scheme/question paper with no examiner's report yet published. Any "the record says..." claim above is scoped to that count.

---

#Sources Used

Primary Pearson documents (all independently verified in research/veridian/WMA14-verified-facts.md, re-applied here to exam-day pacing rather than re-fetched):

• Specification, Issue 3, April 2019 (P4.2.2 assessment structure; Appendix 6 calculator rules).
• Mark schemes: January 2021, June 2022, October 2022, January 2024, October 2024 (Q1), January 2026 (Q3, Q7, Q8, Q9).
• Examiners' Reports (full): January 2021, October 2021, October 2022, January 2024.
• Question papers: June 2022 (9 questions), October 2022 (11 questions), January 2026 (9 questions).

Real questions cited by series and number in this document: Jan 2021 Q3 (general report, proof omission); Oct 2021 Q7(a)+(b) (part (b) not attempted after a struggled part (a)); Oct 2022 Q5 ("wasted much time" on a wrong method); Oct 2024 Q1 (misread capped at B1M1A0A0); Jan 2026 Q7 (16-mark combined question, calculator-technology rubric override), Q8, Q9.

Course files cross-referenced, not duplicated: forge-exam-technique-guide.md (§1's real per-question mark-tariff table, §5's short exam-day list, §6's evidence-base scope note), signal-frequency-and-boundaries.md (§1's topic-frequency table).

Not affiliated with or endorsed by Pearson Edexcel. Every quoted fragment above is under fifteen words with its own series/question attribution; all longer commentary is this course's own, built on — not reproduced from — the cited primary documents.