Level Threshold Guide

#WBS13 Level/Threshold Guide

Provenance — written 2026-09-28; verification pass added 2026-09-29. The 2026-09-29 pass independently re-checked three of this paper's mark schemes directly against their own PDFs (pdftotext -layout, local archive at /Users/neosteinhoff/Documents/SOL/Papers/IAL_Paper3_Complete/ IAL_Materials/Business/) rather than trusting this course's own facts-bank transcriptions for two specific points — the Level 0 floor's wording (§1) and the Level 3/Level 4 wording discrepancy (§3, §4) — and corrected both below where the direct check disagreed with the earlier draft. This document goes one layer deeper than the tariff summary in forge-exam-technique-guide.md §2: rather than stating each level's descriptor once, it works through every real boundary on this paper's three levels-based command words — Discuss (8 marks), Assess (12 marks), Evaluate (20 marks) — and asks what specifically separates a response that stays below a threshold from one that crosses it. Every descriptor quotation below is transcribed from the Jan 2025 mark scheme (WBS13_01_2501_MS) and cross-corroborated, phrase for phrase, against the Oct 2021, Jan 2024, Summer 2024 and Oct 2025 mark schemes, all as recorded in research/veridian/WBS13-verified-facts.md. Every real question, mark, and examiner-report finding cited below is attributed to a specific series and traceable to that same file. Two series cited below — Oct 2025 (WBS13_01_2510_MS, Tesla risk-assessment/workforce-reduction) — were fetched via a third-party mirror because the official qualifications.pearson.com URL could not be located; the facts bank's own provenance note (letterhead, boilerplate and Publications Code independently checked) is the standing caveat on that series, carried forward here rather than repeated at each citation. No verbatim indicative-content bullet list is reproduced in full anywhere below — where the facts bank quotes a full bullet, this document paraphrases the substance and quotes only the short, load-bearing phrase (under ~15 words, attributed).

A note on the "just misses / just clears" examples. Two different kinds of evidence appear below, kept visibly distinct throughout. Where a real, mark-scheme-graded exemplar response exists in the primary sources (a genuine script Pearson's own examiners scored), it is quoted briefly and its real mark is given. Where no borderline real script survives in the material mined — which is most boundaries, since examiner reports quote strong exemplars far more often than mediocre ones — this document builds a short, labelled illustrative pair: two constructed micro-answers to the same real question, written to this course's own understanding of the descriptor language, showing what stays on one side of the line and what crosses it. These illustrative pairs are marked as such. They are not reproductions of any candidate's actual script, and they are not a claim about what mark a real candidate scored — only a concrete demonstration of the qualitative gap the real descriptor wording draws. Nothing below states or implies a raw-mark grade boundary (A*/A/B, etc.); none exists in this course's sources for this paper — see signal-frequency-and-boundaries.md §3.

---

#1. What "levels-based" means on this paper, before the first threshold

Three of this paper's command words are marked against a level ladder rather than a fixed points list: Discuss (8 marks, ladder tops out at Level 3), Assess (12 marks on this paper — the Units-3/4 tariff, not the lighter 10-mark Units-1/2 version — ladder runs to Level 4), and Evaluate (20 marks, ladder runs to Level 4). Define, Calculate, Construct and Explain are point-marked, not levels-based, and are out of scope here (see forge-exam-technique-guide.md §2 for their own traps). Analyse (6 marks) is a genuine, flagged gap: Appendix 6 defines it ("a brief chain of reasoning… does not include evaluation"), but no series read for this course contains a real WBS13 Analyse-tariff question with a level ladder or exemplar attached — this document cannot honestly extend threshold analysis to a tariff nothing in the mined material evidences.

Two facts govern every threshold below, confirmed independently across every examiner report this course has read (Oct 2021, June 2022, Oct 2022, Jan 2023, June 2023, Jan 2025 — 5 of 6 fully mined series carry this exact warning near-verbatim):

1. Marking is holistic, not built from separate Assessment Objectives. A response that attempts evaluation with some context is not guaranteed a top level, and "may only achieve Level 2 if the evaluation is weak" (June 2022 ER, p.3; near-identical wording Oct 2022 ER p.3, Jan 2023 ER p.3, Jan 2025 ER p.4). A threshold is a judgement about the whole response, not a checklist where hitting four of five boxes guarantees a level.
2. A Level 0 floor ("No rewardable material," 0 marks) sits beneath Level 1 on every levels-based question on this paper, on all three tariffs — directly confirmed, not inferred. This course's facts bank had only quoted it explicitly for the 12-mark ladder (Oct 2025 MS, Tesla risk- assessment question). Checking Business_Unit_3_MS_January_2025.pdf directly this pass shows the same explicit "0 No rewardable material" line printed immediately above Level 1 on all three of that paper's ladders: the 8-mark Discuss ladder (Q1c, Spotify sales-forecasting usefulness), the 12-mark Assess ladder (Q1d, Spotify vs Amazon Audible), and the 20-mark Evaluate ladder (Q2, Dyson/SharkNinja PESTLE). The floor is not a tariff-specific feature that happens to be documented for Assess and merely assumed elsewhere — it is one fixed line at the top of every levels-based mark scheme on this paper, regardless of tariff.

The mined material does not explain, beyond this general marking guidance, exactly how a mark is chosen within a level's range (why a Level 3 Assess answer gets 6 rather than 7) — that finer allocation is a genuine gap in what this course's sources can support, and this document does not guess at it. What follows works at the level of the threshold itself: what moves a response from one band into the next.

---

#2. Discuss — 8 marks, caps at Level 3

The ladder (Jan 2025 MS, cross-corroborated against every other series carrying an 8-mark Discuss question read this pass — Oct 2021 IAG/Air Europa, Oct 2022 Pets at Home, Jan 2023 Brompton Bikes, Jan 2025 Spotify):

Appendix 6 fixes the ceiling shape of this tariff before any script is even read: Discuss requires "a brief assessment… no conclusion required." That is why the ladder stops at Level 3 rather than running to a fourth, judgement-bearing level the way Assess and Evaluate do — the top of this tariff is a genuinely two-sided, well-contextualised argument, not a resolved one. The single most repeated examiner finding on this exact tariff, across every series read, is candidates losing the top band not by getting the mechanism wrong but by never supplying the second side at all: Oct 2021's IAG/Air Europa question states plainly that "many candidates ignored the command word 'Discuss' and only gave a one-sided response" (Oct 2021 ER, Q1c).

#2.1 Threshold 0 → 1: rewardable material at all

There is effectively no documented borderline case here in the mined material — a Level 0 script is, by definition, one with nothing creditable in it, and examiner reports understandably do not quote these. The practical marker, inferred directly from Level 1's own wording rather than invented: a single correctly named mechanism, even unapplied to the business in the extract, is "isolated… recall-based" and clears the floor. A response that names no relevant mechanism at all — pure restatement of the question, or content from the wrong topic entirely — has nothing for a Level 1 placement to attach to.

#2.2 Threshold 1 → 2: recall vs application

Real anchor. Jan 2023's Brompton Bikes economies-of-scale question (Q1c, 8-mark Discuss) supplies a genuine, ER-confirmed version of a Level 1 ceiling: weaker candidates were reported "stating or listing different types" of economy of scale, rather than explaining the mechanism of why each one actually lowers Brompton's unit costs (Jan 2023 ER).

Illustrative pair, same question (Brompton Bikes building a new factory; internal economies of scale):

• Stays at Level 1 — "Purchasing economies of scale mean a business can buy in bulk. Managerial economies of scale mean specialist managers." Two terms named, nothing said about Brompton specifically or why either lowers its unit cost — recall, not application.
• Clears into Level 2 — "Building its own factory lets Brompton negotiate steel and component contracts at a much larger volume than its current suppliers offer, spreading the fixed cost of each contract over more units and lowering Brompton's cost per bike." The mechanism is now attached to Brompton's own situation, not stated as a bare definition — but this is still only one side of the argument.

#2.3 Threshold 2 → 3: one good side vs a genuinely balanced one

This is the boundary the examiner reports discuss most, because it is the one candidates cross least often. Real anchor. The same IAG/Air Europa question (Oct 2021 Q1c) had its counter-argument genuinely credited when present: the examiner report confirms marks were given for recognising IAG's takeover disadvantages "might be short lived" given its prior experience — confirming the mark scheme's own indicative content rewards a second side built from evidence already in the extract (IAG already owning Iberia and Vueling), not an invented one.

Illustrative pair, same question (IAG's €1bn takeover of Air Europa):

• Stays at Level 2 — "The takeover could cause diseconomies of scale, since IAG is now a much larger organisation and communication between different parts of the business becomes harder. This could add to IAG's overall costs and demotivate Air Europa's existing staff." Accurate, well- applied, but entirely one-sided — a strong Level 2 answer that never reaches for the other side of the argument at all.
• Clears into Level 3 — the same paragraph, plus: "However, IAG already owns two other Spanish airlines, Iberia and Vueling, so has direct experience managing exactly this kind of integration, and Air Europa's own South American route network may bring in revenue that offsets the extra cost." No conclusion is added — none is required at this tariff — but the second, evidenced side is now present, which is the entire remaining distance to Level 3.

---

#3. Assess — 12 marks on this paper, caps at Level 4

Confirmed empirically across every series read: every Q1(d) and Q1(e) on this paper carries the Units-3/4 Assess tariff (12 marks), never the lighter 10-mark Units-1/2 version. The ladder, cross-corroborated word for word across the Oct 2021 (BA), Jan 2025 (Spotify), and Oct 2025 (Tesla) mark schemes — three non-adjacent series, not one source repeated:

A genuine wording discrepancy — checked directly against three primary-source mark-scheme PDFs this pass, not left as a flagged transcription risk. This document's earlier draft compared Oct 2021's Level 3 wording against Jan 2025's only as transcribed secondhand in this course's facts bank. This pass instead read Business_Unit 3_MS_October_2021.pdf, Business_Unit 3_MS_January_2024.pdf (the Patagonia question) and Business_Unit_3_MS_January_2025.pdf directly, via pdftotext -layout:

• Oct 2021 and Jan 2024 — in both mark schemes, checked independently, Level 3 and Level 4 of the 12-mark ladder open with the identical clause: "Accurate and thorough knowledge and understanding, supported throughout by relevant and effective use of the business behaviour/context." Word for word, no difference between the two levels' opening sentence in either series. The same is true of the 20-mark ladder in both documents (see §4 below).
• Jan 2025 — here, and only here of the three checked directly, Level 3 is genuinely shorter and distinct from Level 4: "Accurate knowledge and understanding, supported by relevant and effective use of the business behaviour/context" (no "and thorough," no "throughout"), while Level 4 keeps the fuller wording. This holds on both the 12-mark and 20-mark ladders in this document too.

Two series checked directly (Oct 2021, Jan 2024) share the fuller wording; the one series checked directly that postdates them (Jan 2025) has the shorter, genuinely-differentiated wording; the facts bank separately records Oct 2025's Tesla mark scheme (mirror-sourced, its own provenance caveat applies, not independently re-checked against a raw PDF this pass) as matching Jan 2025's shorter form. That is a real Pearson wording change, and direct primary-source checking narrows it to somewhere between the January 2024 and January 2025 series specifically — a stronger, more precise claim than "at some point after Oct 2021," which is as far as the previous draft's secondhand comparison could honestly go. Practically, it still changes nothing about where either boundary sits (5–8 marks on the 12-mark ladder, 9–14 on the 20-mark one, either wording) — only how demanding Level 3's own upper edge is described, and, in the pre-2025 wording, how little daylight the mark scheme's own words put between Level 3 and Level 4 at all.

#3.1 Threshold 0 → 1: as with Discuss, undocumented at the boundary itself

No genuine Level 0/1 borderline script survives in the mined examiner reports — unsurprising, since a Level 0 response has nothing worth an examiner quoting. The same inference as §2.1 applies: one correctly named, unapplied mechanism clears the floor into Level 1; nothing relevant to the question does not.

#3.2 Threshold 1 → 2: naming a mechanism vs applying it to the named business

Real anchor. Summer 2024's Bosch SWOT question (Q1d) gives a precisely documented version of this boundary from the opposite direction — candidates who produced a SWOT of Bosch itself, without discussing the benefits of using the tool for developing Bosch's strategy, were reported as "typically placed in the lower levels" (Summer 2024 ER). The mechanism (SWOT itself) was present and even applied to Bosch's own strengths/weaknesses — but the question asked about strategic use of SWOT, not a SWOT of Bosch, so the response stayed low regardless of how thorough the four-quadrant content was.

Illustrative pair, Bosch SWOT question:

• Stays at Level 2 — "Bosch is strong in automotive components and has weaknesses in its complexity as a large multinational. Opportunities exist in EV component demand; threats include new Chinese competitors." A correct SWOT, genuinely applied to Bosch — but it answers "what is Bosch's SWOT," not the question actually asked ("assess the likely benefits… of using a SWOT analysis when developing its business strategy"), so the chain of reasoning stops short of a strategy-development assessment.
• Clears into Level 3 — "By setting its strengths in automotive components against threats from new Chinese competitors, a SWOT analysis lets Bosch identify where its existing capability could defend against the specific threat it faces, rather than reacting generically — informing a strategic decision (for example, whether to defend its current market or diversify into EV components where a strength and an opportunity both point the same way)." The SWOT content is now in service of the actual question — strategy development — not offered as an end in itself.

#3.3 Threshold 2 → 3: developed reasoning, but only one lever named

Real anchor, the sharpest-documented version of this boundary on the whole paper. Two independent real questions — Lush's employee share ownership (June 2023 Q1d) and Five Guys' financial rewards (Oct 2023 Q1d) — both test the exact same trap from different angles, and both examiner reports name it in near-identical language. Five Guys' ER reports candidates who "focused on mainly productivity or motivation rather than labour turnover" were held back, and that reaching the higher levels required the focus to be on labour turnover and its connection to financial rewards specifically. The mark scheme treats the motivation/productivity mechanism and the turnover/retention mechanism as genuinely separate arguments the question specifically asked about the second one.

Illustrative pair, Five Guys financial rewards question ("Assess the likely effectiveness for Five Guys of using financial rewards to reduce labour turnover"):

• Stays at Level 2/entry-Level 3 — "Financial rewards motivate staff to work harder, because money is a strong incentive — Taylor argued this directly. Five Guys' weekly bonus scheme for its best-performing 200 restaurants gives staff a reason to perform well." Accurate, applied, and even uses a real figure — but the mechanism described is productivity, not retention. The question asked about labour turnover specifically, so this answer's chain of reasoning, however developed, is reasoning about the wrong dependent variable.
• Clears into Level 3 — the same factual content, redirected: "Because the bonus is paid weekly and re-earnable, it gives staff a standing, recurring reason to stay at Five Guys rather than leave for a competitor — the reward only pays out to whoever is still there to earn it next week, which is a retention mechanism, not just a productivity one." The figures are identical; what changes is that the reasoning now actually answers the turnover question, which is the entire distance between the two levels here.

A second, independently documented instance of the same boundary: Jan 2024's Patagonia transformative- leadership question found a purely descriptive response was "more likely to achieve level 1 or level 2 at best" — describing what transformative leadership is generically, rather than Patagonia's own use of it, keeps a response below Level 3 regardless of how accurate the description is (Jan 2024 ER).

#3.4 Threshold 3 → 4: a developed, one-sided (or unweighed) case vs a genuinely supported judgement

This is the single most consequential — and most consistently missed — threshold on the paper's 12-mark tariff. Real anchor, confirmed independently across three non-adjacent series (Oct 2021 BA contingency planning, Oct 2023 Five Guys financial rewards, Jan 2024 Patagonia leadership), each examiner report states some close variant of: "a conclusion/judgement is required for 12 mark questions but is not often seen." This is not a formatting instruction — Level 4's own wording is explicit that the judgement has to be supported and has to show "an awareness of competing arguments/factors," not simply asserted after a one-sided case.

Real anchor with a genuine graded script. Peloton's acquisition of Precor (June 2022 Q1d) supplies a real 12/12 exemplar, praised specifically for balancing inorganic growth's speed and market-access advantages against "the cost of the takeover, possible culture clashes and diseconomies of scale" — both sides genuinely weighed, not one asserted and the other merely mentioned.

Illustrative pair, BA's contingency-planning question — "assess the extent to which contingency planning might have reduced" BA's IT-systems-failure risk:

• Stays at Level 3 — "A contingency plan would let BA respond quickly if its systems failed again, reducing the cost of rescheduling flights and the reputational damage seen after the 2017 and 2019 failures. However, contingency planning has an opportunity cost in the time and resources needed to keep the plan tested and current, and BA's own repeated failures show a plan cannot guarantee systems never fail." Both sides are present, developed, and applied — but the paragraph simply stops there. No position is taken on the actual question ("the extent to which"), so the significance of the two competing points is never actually weighed against each other.
• Clears into Level 4 — the same paragraph, plus a closing judgement: "Given that BA's IT failures have recurred despite prior investment, contingency planning is likely to reduce the cost of a failure once it happens more reliably than it reduces the likelihood of one occurring in the first place — so it is a genuine risk-mitigation tool, but only a partial answer to the extent asked about, and should sit alongside direct investment in more resilient IT systems rather than replace it." The facts are unchanged; what closes the gap is a judgement that engages with "the extent to which," conditional on what was actually argued above it — not a flat preference bolted on at the end.

---

#4. Evaluate — 20 marks, caps at Level 4

Both Section B (Q2) and Section C (Q3) are a single 20-mark, source-based Evaluate essay. The ladder (Jan 2025 MS, cross-corroborated against every series carrying a 20-mark Evaluate question read this pass — June 2022 Coca-Cola, June 2023 Apple Watch and Sosyo/Hajoori, Oct 2023 Disney, Summer 2024 Burger King, Oct 2024 Center Parcs, Jan 2025 Dyson/SharkNinja, Oct 2025 Mears Group):

Two features distinguish this ladder from Assess's, both worth naming precisely rather than treating Evaluate as "a bigger Assess." First, Level 3 and Level 4 both explicitly credit using quantitative or qualitative evidence from the source booklet to support a judgement — not merely narrating what the extract says. Second, Level 4's conclusion is asked to go further than Assess's "supported judgement," toward "a solution and/or recommendations" — genuinely prescriptive, not only evaluative.

The same Level 3/Level 4 wording caveat raised in §3 applies here too, directly confirmed rather than assumed. Checking Business_Unit 3_MS_October_2021.pdf and Business_Unit 3_MS_January_2024.pdf directly this pass shows this 20-mark ladder's Level 3 and Level 4 sharing the identical "Accurate and thorough knowledge and understanding, supported throughout…" opening in both series — only from Business_Unit_3_MS_January_2025.pdf onward does Level 3 read as the shorter, genuinely distinct "Accurate knowledge and understanding, supported by relevant…" form quoted in the table above. The mark ranges are unaffected either way; only how sharply Level 3 and Level 4 are worded apart depends on which series.

#4.1 Threshold 0 → 1: as above, undocumented at the boundary itself

No genuine Level 0/1 borderline essay survives in the mined material. The same structural floor applies: a response containing no relevant business content scores 0; the presence of any correctly named, if unapplied, concept clears it into Level 1.

#4.2 Threshold 1 → 2: description of the extract vs analysis using it

Real anchor, the clearest-documented version of this specific boundary anywhere in the facts bank. Oct 2023's Disney resistance-to-change question (Q3) was, by the examiner report's own account, poorly attempted specifically because many candidates were "just paraphrasing the extracts and describing the events" at Disney, as opposed to responses that "really used the extracts to contextualise the potential issues." The report is explicit that the difference is not which side of the argument a candidate opened with, but whether the extract's content was used as evidence for an argument or simply restated (Oct 2023 ER).

Illustrative pair, Disney resistance-to-change question:

• Stays at Level 1/low Level 2 — "Disney has recently appointed a new CEO. The extract explains that some employees are used to the old way of working and some senior managers have left the company since the change was announced." This restates what the source booklet already says, correctly, but connects no cause to any consequence and reaches no argument of its own — description, not evaluation.
• Clears into Level 2/3 — "The departure of senior managers named in the extract is itself a form of resistance — those with the most invested in the old culture are the ones most likely to leave rather than adapt, which risks losing exactly the institutional knowledge Disney needs during a leadership transition. This suggests the scale of resistance is tied to how deeply embedded Disney's prior culture was, not simply to the new CEO's own decisions." The same extract content is now doing analytical work — connecting a named event to a named consequence — rather than being repeated.

#4.3 Threshold 2 → 3: connected reasoning vs reasoning supported by the source's own evidence

Real anchor. June 2023's Apple Watch Porter's-five-forces question (Q2) supplies a genuine 18/20 exemplar that works through three forces in turn — "threat of substitutes, bargaining power of suppliers… and bargaining power of consumers" — each tied to a real figure from the extract (Apple's 33.5% smartwatch share; the 2022 supply-chain disruption), before noting the model's own limitation: Porter's five forces alone "doesn't give us a good idea about the market conditions," so PESTLE should supplement it. What separates this from a Level 2 ceiling is exactly the ladder's own wording: the market-share and supply-chain figures are not narrated, they are used to support each point made.

Illustrative pair, same question:

• Stays at Level 2 — "The threat of substitutes is high because Fitbit, Google and Samsung all make competing smartwatches. This makes Apple's position less secure." A correctly identified force, correctly connected to a consequence — but nothing from the extract itself (market shares, the supply-chain disruption) is used to show how threatening the substitutes actually are.
• Clears into Level 3 — "With Fitbit and Google's joint share at only 8.1% against Apple's 33.5%, the threat of substitutes is real but currently limited in scale — though the 2022 supply-chain disruption shows Apple's own supplier power is a more immediate constraint on maintaining that lead than substitute products are right now." The same forces are named, but the extract's own figures are now doing the evidential work the descriptor rewards, giving the argument something concrete a bare "this makes Apple less secure" does not have.

#4.4 Threshold 3 → 4: a conclusion vs an effective one that recommends something

Real anchor. Summer 2024's Burger King CPA question (Q3) supplies a genuine 18/20 script that correctly calculated the real 11-week project duration and evaluated CPA specifically applied to Burger King's renovation programme — but lost marks specifically on "an underdeveloped conclusion," not on the calculation or the analysis (Summer 2024 ER). This is a rare, real, documented case of a script that clears every earlier threshold cleanly and is still held below full marks by the Level 3→4 gap specifically. June 2022's Coca-Cola CPA question, by contrast, supplies a genuine 20/20 exemplar for the same command word and similar content — weighing CPA's planning benefit against it being "based on past data" and ignoring external delays, before reaching a conclusion the mark scheme credited as complete.

Illustrative pair, Burger King CPA question ("evaluate the benefits to Burger King of using CPA for its renovation programme"):

• Stays at Level 3, upper end — a fully worked 11-week network diagram, correct float identification on the non-critical activity, and a balanced case (CPA reduces the project's own risk and cost by forcing an activity-by-activity assessment of requirements, against CPA's own reliance on a single site's layout not generalising across Burger King's whole rollout) — ending with: "CPA is therefore a genuinely useful planning tool for this renovation programme." The judgement is present but generic — it restates that CPA is useful without saying how confidently, or what it would still leave Burger King needing to manage.
• Clears into Level 4 — the same worked network and balanced case, ending instead with: "CPA is most valuable for planning a single renovation's own timeline and resource allocation, but since different restaurants have different layouts and local contractor availability, Burger King should treat each site's network as a fresh calculation rather than assuming this one generalises across the whole rollout — recommending CPA be paired with a standardised pre-renovation site survey so each new network starts from comparable, reliable estimates." The evidence and analysis are unchanged from the Level-3 version; what is added is a conclusion that proposes something concrete, matching Level 4's own "solution and/or recommendations" wording rather than a flat restatement of usefulness.

A second, independently documented instance of the identical threshold: June 2023's Sosyo/Hajoori franchising question (Q3) supplies a genuine 17/20 exemplar explicitly credited for evaluating franchising fully and then proposing a genuine alternative (social-media-led expansion) — but the mark scheme is precise that the alternative only earns credit provided franchising itself is still fully evaluated first, not displaced by it. A conclusion that swaps to the alternative instead of completing the evaluation actually asked for does not clear this same boundary, however well-argued the alternative itself is.

---

#5. Three patterns that repeat across all three tariffs

Having worked through eleven boundaries individually, three findings recur often enough, across independent series and independent topics, to be worth stating once as cross-cutting rather than re-derived at every threshold:

1. One-sidedness caps a response below the top band on every levels-based tariff on this paper, not just Discuss (where it is the headline trap). Discuss's Level 3 needs a second side with no conclusion; Assess's and Evaluate's Level 4 need a second side and a judgement that weighs it. The single side offered is very often accurate and well-developed on both halves of the ladder — the ceiling is about coverage of the argument, not the quality of any one part of it.
2. Copying or narrating the source material, rather than using it as evidence, caps a response low regardless of tariff. Confirmed independently on a 12-mark question (Oct 2021 BA: "many candidates simply copied large chunks of the information from the Extracts") and a 20-mark one (Oct 2023 Disney: "just copied out the extract and described events"). This is a distinct trap from one-sidedness — a narrated extract can technically cover both sides of an issue and still stay low, because nothing has been done with the material.
3. Breadth of coverage is not the same lever as depth of argument. Summer 2024's Bosch ER states this explicitly: "it was not necessary to cover all four aspects of SWOT to achieve level 4." A response naming every possible factor, mechanism, or quadrant a topic offers does not automatically outscore one that develops fewer points further — the ladder rewards developed reasoning and a supported judgement, not exhaustiveness.

---

#6. What this document cannot tell you

Named directly, not implied: this paper's sources give no raw-mark-to-grade conversion for WBS13 at any series (see signal-frequency-and-boundaries.md §3). "Threshold" throughout this document means a boundary between two fixed Level mark-bands, never a raw score needed for a specific grade — those two concepts are easy to conflate and this course's sources only support the first. They give no worked account of how a mark is chosen within a level's range beyond the general marking guidance already described in §1. They give no level ladder at all for the Analyse tariff, because no WBS13 series read for this course contains one. And the 0→1 threshold on every tariff is inferred from Level 1's own wording rather than drawn from a documented borderline script, because examiner reports do not, in the material mined, quote responses that scored nothing. Where evidence runs out, this document has said so rather than filling the gap.