Level Threshold Guide

Provenance: compiled 2026-09-29. Primary source is research/veridian/WMA11-verified-facts.md §3 (mark-scheme conventions, verified verbatim against the January 2023 and January 2024 mark-scheme PDFs) and §4 (per-topic examiner-report material, itself citing named, dated primary-source PDFs — question papers, mark schemes and examiner reports — for each series quoted). Every question this document builds a ladder from is a real, verified WMA11 question; no ladder, mark line or examiner quote below was invented or reconstructed from a general sense of how Edexcel mark schemes "usually" work. Cross-checked against every existing document in content/veridian/technique-hub/wma11/ and every lesson in content/veridian/curriculum/wma11/ before writing, specifically to avoid restating material those documents already cover — see §0 and §4 below for exactly what was found and excluded. Not affiliated with or endorsed by Pearson Edexcel.

#WMA11 — Level/Threshold Guide

#Pearson Edexcel IAL Mathematics, Unit 1: Pure Mathematics P1 — mark-band boundaries, by real example

#0. What this document is, and an honest note on what it found

A Maths paper has no essay-style "Level 3 vs Level 4" descriptor ladder — WMA11's only real level system is the M / A / B / dM / ddM mark-type code, applied one discrete point at a time. So a genuine "threshold guide" for this paper has to work at that grain: for a specific real question, exactly what document, on the page, moves a script from one mark to the next, and what real, examiner-confirmed response sits on each side of that line. That is what this document tries to do — not repeat the tariff-and-technique material forge-exam-technique-guide.md already covers (quoting the formula first, the three-route menu for solving a quadratic, the per-term differentiation/integration mark, the dM/ddM chain on Oct 2021 Q2, the abbreviation list), and not repeat the frequency/strand analysis signal-frequency-and-boundaries.md already covers.

The honest finding this pass has to report before anything else: this course's own prior sessions have already mined almost every rich, multi-mark M/A/B breakdown in the facts bank into a specific lesson, CODEX deep-dive, or SIGNAL topic brief, several of them built explicitly around a "just misses vs just clears" contrast at the exact mark boundary this document exists to document. Checked directly, not assumed: Jan 2019 Q7(a)'s sine-rule ambiguous-case ladder is already built into sine-and-cosine-rules-ambiguous-case.ts with "Earned"/"Lost" framing at each mark; Jan 2023 Q2's perpendicularity-proof A1 (the mark lost for a correct calculation with no closing sentence) is already built into perpendicularity-proofs-and-coordinate-geometry.ts the same way; June 2022 Q5(a)'s "1 out of 4 was a very common score" finding is already a full marked-solution in quadratic-functions.ts; Jan 2023 Q3's split-then-integrate dM1 structure is already in signal-topic-brief-splitting-fractions-before-integrating.md; Oct 2021 Q2's dM1/ddM1 chain is already in paper-anatomy.ts, forge-exam-technique-guide.md §2.5 and codex-formula-and-method-bank.md. Per this course's own standing rule against padding, this document does not rebuild any of those — §4 below names exactly which document to open for each one instead.

What was left, once that material was set aside, is smaller but genuinely new: two documented special-case (SC) mark rulings that no existing WMA11 document in this repo currently ladders — one already used only as a one-line abbreviation example in FORGE, one not used anywhere at all — plus a real, verbatim contrast in the general marking rules that no existing document draws out explicitly. §1 covers the general rules as thresholds in their own right; §2 builds the two fresh ladders in full; §3 states, honestly, two general rules this paper's marking guidance states verbatim but for which the verified corpus contains no concrete WMA11 worked instance — flagged rather than invented.

---

#1. The general rules, read as thresholds rather than as advice

forge-exam-technique-guide.md §2.6 already lists Rule 4 and Rule 5 as things to know. Read side by side, they draw a real boundary line this course has not stated explicitly before: a wrong number does not cost the same amount depending on why it's wrong.

• A genuine misread (the candidate copies a value from the question incorrectly, but the question itself is not made easier or changed in character by the error) costs a flat deduction of two from the A or B marks in the affected part — Rule 5, verbatim: "deduct two from any A or B marks gained." The candidate keeps every M mark, and keeps A/B marks above the two-mark deduction.
• A genuine wrong-working error that is not a misread has no such floor. Because Rule 4 states A marks are cao "unless shown... as A1 ft," a wrong value produced by an error in the method itself is not owed any follow-through credit by default — the marker follows it through only where the mark scheme explicitly says ft applies, and stops entirely at a "manifestly absurd" answer (Rule 4's own closing phrase).

The practical threshold this produces: the two-mark floor is a misread-specific protection, not a general one. A candidate who copies "6" for "9" from the question paper and works flawlessly from there is capped at losing two marks in that part. A candidate who copies the question correctly but makes an algebraic slip on the same line has no such cap — the mark scheme's default is to award nothing further from that point unless a specific ft note says otherwise. Two errors that look equally "small" on the page can cost very different amounts, and which rule applies depends on whether the wrong number came from misreading the question or from a mistake made while working with it — not on how far downstream the consequences run.

No concrete WMA11 question in the verified corpus illustrates this exact contrast with real numbers on both sides (see §3 for why this is stated as a gap rather than papered over with an invented example). The rule itself, however, is verbatim and verified against two independent series' mark schemes (facts bank §3), and is a real threshold a student should know exists before it costs them marks on a live paper.

---

#2. Two threshold ladders not yet built anywhere else in this course

#2.1 Oct 2021 Q3(i) — the B1-only floor on a reciprocal inequality

The question (3 marks, verified directly from the primary-source mark scheme and examiner report, physicsandmathstutor.com mirror October 2021 MS.pdf and Pearson's own wma11-01-pef-20220113.pdf): solve 3/x > 4.

The real ladder, verbatim mark-scheme structure:

The floor that makes this ladder worth documenting: the mark scheme records a special case sitting exactly at B1 with nothing above it — a script that rearranges the inequality the "naive" way (treating x as positive throughout, without the ×x² step) and states x < 3/4 only. That response never reaches the M1 region-choice step at all, because it never produces the second critical value (x = 0) that a case-correct method requires — so it scores B1 and stops, 1 of 3. The examiner report confirms this was not a rare edge case but the single most common wrong answer on the part, produced by candidates who rearranged the inequality directly rather than clearing the fraction by multiplying through by x² (a technique the mark scheme's own top-scoring route uses precisely because x² ≥ 0 always, so it never needs a separate case split on the sign of x).

Just misses vs just clears, concretely:

• Just misses (B1, 1/3): "3/x > 4 ⇒ 3 > 4x ⇒ x < 3/4." One correct-looking critical value, no second one, no case reasoning shown — scores the bottom rung and stops there, because the region-choice M1 has nothing to act on without both critical values.
• Just clears (B1 M1 A1, 3/3): "3/x > 4 ⇒ 3x > 4x² [multiplying by x², always ≥ 0] ⇒ 4x² − 3x < 0 ⇒ x(4x−3) < 0 ⇒ critical values 0, 3/4 ⇒ 0 < x < 3/4." Same final-looking fragment ("3/4") appears in both scripts — the ladder is decided by whether the second critical value and the region reasoning are visible, not by whether "3/4" appears on the page.

#2.2 June 2023 Q2(c) — the M-marks-without-A-marks floor on a word-problem quadratic

The question (6 marks total across three parts; verified directly against WMA11_01_2306_MS, QP Log Number P72868RRA): a rectangular pitch, length x, width y, x > y, perimeter 350 m, area 7350 m². Parts (a)–(b) ask for the two equations (2x+2y=350, xy=7350; B1 each). Part (c), 4 marks, asks for the values of x and y.

The real ladder for part (c):

The floor this question's mark scheme documents, and no existing WMA11 document in this repo yet uses: a script that reaches the correct pair of numbers by trial-and-improvement, or by guessing, with no quadratic ever formed or solved on the page, is capped at M1 M1, A0 A0 — two of the four marks, not four, however correct the final x/y pair is. The two method marks are awarded because the mark scheme treats "correctly identifying that 70 and 105 satisfy both conditions" as evidence of the same reasoning the algebraic route would show; the two accuracy marks are withheld because they are gated on the quadratic itself having been formed and solved, not merely on the final numbers being right.

Just misses vs just clears, concretely:

• Just misses (M1 M1 A0 A0, 2/4): "Try x = 100, y = 75: perimeter 350 ✓, area 7500 ✗. Try x = 105, y = 70: perimeter 350 ✓, area 7350 ✓. So x = 105, y = 70." Ends at the correct, correctly-assigned pair — and still caps at half marks, because no quadratic was ever formed.
• Just clears (M1 dM1 A1 A1, 4/4): "y = 175 − x. x(175 − x) = 7350 ⇒ x² − 175x + 7350 = 0 ⇒ (x−70)(x−105) = 0 ⇒ x = 70 or 105. Since x > y, x = 105, y = 70." Reaches the identical final pair — the two extra marks come entirely from the algebraic route being visible, not from the answer being any more correct.

This is the sharpest real illustration in this paper's verified corpus of a general truth the mark scheme states abstractly in its General Principles (facts bank §3) but rarely shows this cleanly with real numbers on both sides: on a paper marked M/A/B, a correct final answer with no visible method is not "nearly full marks with a presentation penalty" — it is capped at whatever the method marks alone are worth, and can be exactly half the total even when nothing about the final answer is wrong.

---

#3. Two rules this paper states verbatim, with no verified worked instance — flagged, not invented

Per this course's standing rule against fabricating material to fill a gap, these are named honestly rather than illustrated with an invented example:

• cso ("correct solution only") vs cao. The mark-scheme conventions (facts bank §3) define cso as stricter than cao: "There must be no errors in this part of the question... to obtain this mark" — meaning a cso-coded mark can be lost even when the final numeric answer is correct, if an earlier line in that same part contained an error that happened to cancel out. This is a real, verbatim WMA11 marking convention. No question in the verified corpus for this course, across the nine series touched, has yet been read closely enough to confirm a specific instance of a WMA11 cso mark being lost this way. A future primary-source pass targeting mark schemes for a cso-coded line specifically (the abbreviation appears in the general conventions but was not the object of any of the topic-focused extractions this course's research pass has done so far) would be needed before this threshold could be shown with a real example instead of stated as a rule.
• Rule 4's "manifestly absurd" cutoff. Rule 4 states follow-through marks are withheld from a "manifestly absurd" answer even after a genuine misread, but the verified corpus contains no example of a real WMA11 script or mark-scheme annotation that names a specific answer as manifestly absurd. The rule is real and verbatim; a concrete WMA11 illustration of where that line actually falls is not yet part of this course's verified material.

---

#4. Where the rest of this paper's threshold material already lives

Every other multi-mark ladder rich enough to build a "just misses vs just clears" pair from was checked directly against this repo before writing this document, and found already built out in one of the following:

This document ends here rather than re-describing any row in that table in its own words. What it adds — §1's misread-vs-error contrast, and §2's two special-case ladders — is the genuine remainder once everything already covered is set aside.