Judgements Guide

Provenance (checked 2026-09-29): this document verifies, and rebuilds, a raw source file — ~/Downloads/FINSTO/THE DEFINITIVE GUIDE TO JUDGEMENTS — WBS12.md — which claimed to be a technique guide for the supported-judgement/effective-conclusion element of WBS12's Assess and Evaluate questions. That source was NOT trusted on sight. It was checked against research/veridian/WBS12-verified-facts.md (read in full) and then, for every claim not already covered there, against real primary-source PDFs: the October 2023 WBS12/01 question paper, mark scheme and examiner report, extracted directly with pdftotext -layout from ~/Documents/SOL/Old/P2 Materials/Business/October_2023/ (this is the series the source's central worked example — Amazon's Kindle withdrawal from China — is drawn from; note the equivalent October_2023 folder does not exist in the newer ~/Downloads/PAPERI/... archive, so the older SOL archive was used instead), plus the January 2024 examiner report for a second Assess-tariff data point. content/veridian/technique-hub/wbs12/forge-exam-technique-guide.md was read in full first so this file stays additive to it rather than repeating its tariff-by-tariff material. Quotations from Pearson material are kept short and attributed by series/question rather than reproduced at length, per this project's copyright convention.

#WBS12 Judgements Guide — What a Supported Judgement Actually Needs

forge-exam-technique-guide.md already establishes that Assess (10 marks) needs "a supported judgement" and Evaluate (20 marks) needs "an effective conclusion that proposes a solution and/or recommendations" to reach Level 4, and that an unconditional conclusion caps the level (external-influences.ts's existing unconditional-conclusion-caps-the-level trap). This document goes one level deeper: what, specifically, does a conclusion need to contain to satisfy that Level 4 language — verified against what Pearson's mark schemes and examiner reports actually say, not against a framework invented and dressed up as if it were Pearson's own wording.

The headline finding of the verification pass: the source file's core technique content — decide, justify, anchor to the extract, state a condition, state a counter-condition — is a reasonable and mostly defensible framework, and several of its specific factual claims about the Amazon/Kindle question check out exactly against the real extract and mark scheme. But the source's single most-repeated worked mechanism — that "Chinese government regulation... forced Airbnb and LinkedIn to exit" China, used as the load-bearing condition in nearly every worked judgement example in the source — does not appear anywhere in the primary source. The real extract gives no reason for Airbnb's or LinkedIn's exit at all, and the real mark scheme files that example under "market conditions," not "government regulation." That correction, and what it implies for how a condition should actually be built, is covered in §3 below.

---

#1. What the mark scheme actually says, verbatim, and what it doesn't

Two sentences carry all the weight in this area. Both were checked directly against the October 2023 mark scheme PDF (identical wording confirmed in January 2024's, matching the verified-facts file's own record of this being stable phrasing across series):

• 10-mark Assess, Level 4 (8–10 marks): "...shows an awareness of competing arguments/factors, leading to a supported judgement."
• 20-mark Evaluate, Level 4 (15–20 marks): "...leading to balanced comparisons, judgements and an effective conclusion that proposes a solution and/or recommendations."

Two things the source file got right about these sentences: the two tariffs genuinely ask for different things (Assess wants a decided-between judgement; Evaluate additionally wants a solution or recommendation), and the phrase "an effective conclusion is sought" is real, cited correctly in the source's opening line. It is real — but the source's framing ("the examiner report says the same thing every single series") overstates it. The verified-facts file's own cross-series check found that exact phrase from October 2021 onward (recurring in January 2022, October 2022, January 2023, October 2023, January 2024, October 2024 — 7 of the 13 fully-readable reports, not all 13, and not the two 2019/2020 reports before it). Checked independently for this pass: it is present, worded identically, in both the October 2023 and January 2024 reports' Question 3 sections. Treat it as a standing, current phrase since 2021 — not as evidence it has always been the rule.

Nowhere in the mark scheme, level descriptors, or the two examiner reports checked for this pass is there a formal, numbered breakdown of a judgement into named "elements." That structure — decision, justification, extract anchor, condition, counter-condition, new addition — is the source file's own analytical framework, not a Pearson taxonomy. It is a defensible way of unpacking what "awareness of the significance of competing arguments... leading to a supported judgement" actually requires in practice, and it is kept below because it holds up against the verified examples — but it should not be presented to a student as Pearson's own wording, and this rewrite does not present it that way.

---

#2. The judgement, rebuilt from verified pieces

A judgement earns Level 4 credit by doing four things the body of the answer does not do on its own:

1. It decides. Not "both sides have merit" — a stated preference, in the question's own terms. The October 2023 examiner report's Question 3 commentary (Amazon Kindle) confirms the opposite failure mode directly: candidates who struggled were the ones who "struggled to apply the extracts appropriately or provide balanced arguments" and who showed they "lacked understanding of internal and external causes" — not candidates who over-committed. There is no verified quote, in the report checked for this pass, of examiners specifically penalising decisiveness or rewarding hedging; the safer reading of the evidence is that a stated, reasoned decision is asked for and an unbalanced or generic one is what gets marked down — not that decisiveness itself is risky.

2. It weighs, not just restates. The genuinely verified mechanism here is the report's own language about why higher-level answers on Q3 succeeded: "higher level attainment came from an awareness of competing arguments such as contrasting the likely impact of internal causes with external causes but also, by evaluating the likelihood of the causes analysed" (Oct 2023 examiner report, Q3). Contrasting and evaluating likelihood is the comparative move — restating one side's impact a second time in the conclusion is not that move, because it doesn't compare anything.

3. It uses the extract, specifically. Every level descriptor above Level 2 on this paper requires "quantitative and/or qualitative information" — and the report is explicit that this has to be genuine application, not extract text lifted into a stand-alone sentence: "Rewriting the extracts to state the information provided to candidates in the first place, rather than answering the question, did not enable the candidate to progress through the levels" (Oct 2023 Q3). A judgement that names a specific figure or detail and explains what it implies for the decision is different from a judgement that quotes the extract without using it.

4. It states a condition — a real one, falsifiable within the scenario. This is the part external-influences.ts's existing trap already flags as an unconditional-conclusion cap, and it's the part the source file's central worked example gets structurally right (a judgement should say what it depends on) even where its factual content for the Amazon example is wrong (§3). A condition earns credit by being specific to this business's own extract and capable of being false — not a vague "this depends on market conditions" that could be bolted onto any answer on the paper.

None of this requires the source file's precise mark-by-mark table ("16/20 = elements 1+4 present... 19/20 = all five elements plus a new framing..."). No mark scheme or examiner report checked for this pass gives sub-band guidance at that resolution — Pearson's own descriptors are written at the level-band level (e.g. 15–20 as one single Level 4 band with one descriptor), and inventing a fake five-mark ladder inside it is exactly the kind of unearned precision this verification pass exists to catch. Where the mark scheme says "the response gets marked toward the top or bottom of whichever level it best matches" (the general-marking-guidance mechanism forge-exam-technique-guide.md already documents from the January 2025 report), that's the real mechanism — not a per-element points table.

---

#3. The corrected worked example: Amazon's Kindle withdrawal from China (October 2023, Q3, 20 marks)

The real question, verified verbatim against the mark scheme: "Evaluate whether Amazon's decision to stop selling its Kindle in China was due to internal or external causes of business failure." This is genuinely Pearson's Section C question that series (worth 20 of the paper's 80 marks), and the source file's choice of this as its central example is sound — but several of its specific "extract" details needed checking line by line against the real Extract E and Extract F. Here is what verified, what didn't, and what a corrected judgement should actually say.

#What checked out exactly

• The 10-year decline. Extract F states, close to verbatim: the Kindle "did not suddenly fail in China but declined slowly over 10 years" as the Chinese e-book market "changed and evolved" while the Kindle did not. This is real, and it is genuinely strong extract-anchor material for a judgement — a slow, decade-long decline is a different shape of evidence than a sudden shock, and a judgement can legitimately use that shape to argue for an internal, complacency-driven explanation over an external shock explanation.
• CITIC Press Group and Commercial Press. Real, verbatim in Extract F — Amazon's "good relationship with Chinese publishing houses, such as the CITIC Press Group and Commercial Press" is named as one of the Kindle's early selling points, which the mark scheme's own indicative content also uses to argue supplier problems were unlikely to be the cause.
• The Harry Potter detail. Real, verbatim in Extract F: the Kindle "neglected online fiction, with popular titles such as the 'Harry Potter' series either incomplete or missing altogether." The source file's use of this as evidence of an internal marketing/content failure is faithful to what the extract actually says.
• Colour e-ink screens — for comics, not manga. Extract F names competitor devices with "colour e-ink screens for reading comics." The source file's recommendation invents a "manga market" that the extract does not mention — a small but real embellishment, corrected below.
• Xiaomi, iFlytek and Huawei; the 2016 market-leader phase; the 500 million figure; the 18%/$6bn figure. All verified, verbatim or near-verbatim, in Extract E. One nuance worth stating precisely where the source blurred it: the "$6bn, up 18%" figure describes the size and growth of China's whole digital-reading market in 2021, not Amazon's own Kindle revenue specifically — the mark scheme's own point from this is that "even a lower market share may still have been a significant revenue for Amazon," i.e. it's evidence the overall market was growing (an external condition favourable to any e-reader), not evidence about Amazon's own sales trend.
• "An effective conclusion is sought" for this exact question. Verified verbatim in the October 2023 report's own Q3 section, closing line.

#What did not check out — and needs to be dropped

The regulation mechanism. The source file's central condition and counter-condition, repeated across nearly every worked example in it, is some version of: "This holds only if Chinese government regulation did not restrict Amazon's operations — as it did with Airbnb and LinkedIn (Extract E)." Checked directly against the real Extract E: it names Airbnb and "Microsoft's LinkedIn" as having left or reduced China operations, and Amazon's own store closure — but gives no stated reason for any of them. The extract's own words: "Amazon did not give a reason for leaving the Chinese market." The mark scheme's indicative content independently confirms this is filed under "market conditions" as an external cause (alongside competition, supplier problems and government regulation as separate, distinct categories) — not attributed to regulation. There is no sentence anywhere in the extract, mark scheme or examiner report checked for this pass that ties Airbnb's or LinkedIn's exit to Chinese government action. This is not a small wording slip: it is the mechanism the source file's Element 4 (condition) and Element 5 (counter-condition) run on in nearly every version of the worked example, and it does not check out. Any judgement built on "this holds unless regulation forced Amazon out too, as it did Airbnb and LinkedIn" is asserting a causal claim the extract does not make and should not be written in an exam answer as if it were extract-confirmed.

This is worth dwelling on because it's a genuinely useful finding about why the fabrication happened, not just that it did: real-world background knowledge (LinkedIn's well-publicised 2021 China exit is often described in general reporting as regulation-driven) got imported into the answer and dressed up with a specific extract letter ("Extract E") as if the exam text said so. That is exactly the failure mode a business-extract answer has to avoid — and exactly the failure mode this whole verification task exists to catch before it reaches a student's revision material.

What a genuine, falsifiable condition looks like instead, built only from what the extract actually supports:

"This judgement — that internal complacency was the primary cause — holds only if Amazon's own product and marketing failures, not a market-wide retreat, were the decisive factor. Extract E lists Amazon's exit alongside Airbnb's and LinkedIn's without giving a reason for any of them; if whatever caused those two departures also applied to Amazon's Kindle business specifically, the slow ten-year decline would be better read as one company's slice of an industry-wide external pressure rather than evidence of internal drift, and external causes would carry more weight than this judgement allows."

This version is honestly conditional — it names the real gap in the evidence (no stated reason, for any of the three companies) rather than inventing a mechanism to fill it, and it is falsifiable in the way Element 4 is supposed to be: if the premise turned out true, the conclusion would genuinely have to change.

A second, independently useful fact the source file never mentions: the mark scheme's own indicative content includes a bullet explicitly warning that a single cause can be legitimately argued as both internal and external at once — "consumer trends change over time... this could be seen as both an internal and an external cause of failure for Amazon." A Level 4 judgement that acknowledges this ambiguity, rather than forcing every fact into a clean internal/external bucket, is doing exactly the "awareness of the validity and significance of competing arguments" the descriptor asks for — arguably better than a judgement that pretends the classification is clean. This is a genuinely new, verified finding worth building into technique: the strongest judgement on this exact question type may be the one that names which piece of evidence resists the internal/external split, not the one that argues hardest for one side.

#A corrected, fully verified judgement for this question

Decision: On balance, internal causes were the more significant driver of the Kindle's withdrawal from China. Weighing, extract-anchored: Amazon's own Kindle did not fail suddenly — Extract F is explicit that it "declined slowly over 10 years" while the Chinese e-book market around it changed. A sudden, market-wide shock produces a sudden collapse; a decade-long fade while Amazon still held early advantages — the CITIC Press Group and Commercial Press relationships, an early lead from 2016 as China's largest Kindle market — points more toward accumulated internal neglect (missing colour e-ink and comics support, an incomplete Harry Potter catalogue) than toward a single external cause. Condition: This holds only if Amazon's specific product and marketing failures, not the market-wide retreat Extract E also documents for Airbnb and LinkedIn, were the decisive factor — and the extract does not say what caused those two departures, which is a genuine limit on how confident this judgement can be. What's new: Notably, the mark scheme's own indicative content flags that a factor like changing consumer trends can be read as both an internal and an external cause simultaneously — this judgement's own internal/external split is therefore itself a simplification the evidence only partly supports, which is worth naming rather than hiding.

That last sentence is the kind of self-aware, evidence-led qualification the descriptor's "full awareness of the validity and significance of competing arguments" is actually asking for — built from a real mark-scheme bullet, not an invented one.

---

#4. Two shorter, fully verified illustrations at the Discuss and Assess tariffs

The source file's shorter worked examples (Arditi Tours, a bus company; Grupo Tamazula, a Mexican hot-sauce producer — both real businesses from the same October 2023 paper, Sections A and B) were checked the same way. Both check out well, with one correction.

Q1(d), 8-mark Discuss — margin of safety, Arditi Tours (no conclusion required at this tariff, but a strong two-sided answer still benefits from stating what it depends on). Verified directly against Extract A and the October 2023 examiner report: the break-even point that April was 17 passengers per journey; the actual average across the day's four departures (15, 21, 29, 11) gives a margin of safety of 8 passengers over the whole day; diesel was €1.58 a litre and rising; and the extract states plainly that "April is a month of average demand for seats." The examiner report's own commentary on this question confirms the strongest answers did exactly this: calculated the 8-passenger margin, called it "enough to be reasonably comfortable based on the April figures," then countered it with the rising diesel cost and fluctuating demand. A genuinely verified two-sided answer:

"Arditi Tours has a reasonably comfortable 8-passenger margin of safety across the day in April — a month the extract itself describes as average demand, meaning this isn't a best-case reading. However, diesel — at €1.58 a litre and rising — is pushing up costs, and demand data shows real volatility between journeys (from 11 passengers on the 15:00 service to 29 on the 11:00), so the comfortable daily average conceals genuinely loss-making individual services."

Q2(d), Discuss the benefits of being a plc — Grupo Tamazula. Verified against the mark scheme and examiner report: raising share capital to keep the 26,000m² factory's equipment modern, and a "more prestigious profile" helping against a "high number of competitors," are both real, credited points — and so is the counter-argument the source file used: outside shareholder control could threaten "the traditions from its formation in 1960 and its products being made from the finest ingredients." The examiner report separately confirms the stronger answers were the ones linking share capital specifically to "the purchase of modern equipment and/or fund research regarding the US and Canadian export markets" — a real, extract-anchored causal chain, not a generic "plcs can raise more money" assertion.

One correction to the source's Q1(e) recommendation: the source suggests "offering Wi-Fi on longer journeys" as an alternative to a price rise. The real mark scheme's credited alternative is different — "offering more services such as TV on journeys" — alongside charging separately for the two pieces of luggage the ticket price currently includes. Wi-Fi is a plausible real-world equivalent, but it is not what the extract or mark scheme actually says, and a student citing it as if it were extract-supported would be making the same category of error flagged in §3.

---

#5. Verification ledger

---

#6. Honest summary of the verification pass

Of the source file's specific, checkable factual claims (extract details, mark-scheme wording, examiner-report quotes — excluding its own invented framework labels and mark-by-mark predictions, which are not factual claims to verify but presentation choices), roughly two-thirds checked out against the primary sources, most of them exactly. The failures cluster in one place: the regulation/Airbnb/LinkedIn causal mechanism, which is wrong on both the cause (no reason is given in the extract) and the categorisation (the mark scheme calls it "market conditions," not regulation) — and which the source file used as the load-bearing condition in nearly every one of its worked judgement examples, not as an isolated slip. That single fabricated mechanism accounts for most of the failed claims by volume, since it recurs across the source's Elements 4 and 5 in multiple worked examples rather than appearing once.

Two claims should be treated as genuinely new findings from this pass, not present in the source at all: the mark scheme's own acknowledgement that a cause can be legitimately both internal and external, and the extract's explicit statement that Amazon gave no reason for leaving China — both of which make for a more sophisticated, better-qualified judgement than anything in the original source, once you build them in honestly.