Error Pattern Tracker

FORGE · WEC14 Error Pattern Tracker — compiled 2026-09-29. Every category and citation below is drawn directly from research/veridian/WEC14-verified-facts.md's own verification against real Pearson mark schemes and examiner reports (14 exam series, October 2020–January 2025, 13 with a matching examiner report, plus one narrower June 2025 check — see that file's "Series coverage log" for the exact denominator). No error pattern below is invented, estimated, or given an invented likelihood: each one is either a direct examiner-reported mistake, tied to a named series and question, or a mark-scheme mechanic ("N.B." gates) verified against the raw PDF. Not affiliated with or endorsed by Pearson Edexcel.

#WEC14 Error Pattern Tracker (Unit 4: Developments in the Global Economy)

#What this is

This paper's technique-hub already has a technique layer (forge-exam-technique-guide.md, tariff-by-tariff), a frequency/boundaries layer (signal-frequency-and-boundaries.md), a definitions/KAA bank (codex-definitions-and-application-bank.md), six per-topic deep dives and briefs, and the curriculum lessons themselves (each already carrying its own worked chains and trap-taxonomy). None of those is a place to log your own mistakes across multiple timed attempts and watch which ones repeat. This document is that place: Part 1 turns the archive's confirmed, examiner-evidenced error patterns into a fixed set of category codes; Part 2 is a blank log to fill in after every timed Section A/B/C attempt; Part 3 is a running tally so a pattern shows up as a number, not a feeling; Part 4 is what to actually do about a code once it recurs.

This document does not re-explain the economics behind any trap — where a curriculum lesson already has the full worked mechanism (diagram, worked chain, trap-taxonomy entry), the category below names it and points there rather than duplicating it.

---

#Part 1 — The confirmed error-pattern categories

Ten categories, each evidenced by a named series and question (or, where the source explicitly states one, a real denominator like "5 of 13 series"). These are not the only mistakes possible on this paper — they are the ones the real archive actually documents recurring. If your own log surfaces a mistake that doesn't fit any of these ten, add it as its own row in Part 2 anyway; do not force it into the nearest code.

#EP1 — Percentage vs percentage-point confusion

Spec ref: cross-cutting (QS2, spec Appendix 7 p. 69); hits 4.3.2.3 terms-of-trade calculations hardest, but recurs on interest-rate and tax-rate questions too.

Mechanism: getting the arithmetic right, then losing the mark anyway by labelling the answer with the wrong unit — writing "X%" when the question's change is a percentage-point change, or vice versa.

Evidence:

• October 2020 Q7(a) (terms-of-trade index calculation): candidates who calculated correctly still lost the application mark by writing "22.22%" instead of "22.22 percentage points" — the examiner report's standard phrasing is that candidates must "use the data carefully for calculation-based questions." Repeated pattern flagged again in January 2023 and October 2023.
• October 2020 Q7(d) (base rate 4% → 3.5%): candidates who wrote "a 0.5% fall" instead of "a 0.5 percentage point fall" could not access the application mark for it.
• January 2022 Q4 MCQ (income tax 45% → 47%): examiner report states candidates should be aware of "the difference between percentage change and percentage point change."

Curriculum pointer: terms-of-trade-and-trading-blocs.ts's percent-vs-percentage-point-on-a-tot-calculation trap-taxonomy entry has the worked mechanism.

#EP2 — Paired-definition reversal

Spec ref: 4.3.4.1 (absolute vs relative poverty); 4.3.3.1 (current-account deficit vs surplus); 4.3.3.2(d)/(e) (devaluation/revaluation vs depreciation/appreciation).

Mechanism: the question asks for one half of a defined opposite pair, and the answer either explains the other half, or explains both halves backwards.

Evidence:

• January 2021 Q7(c) (relative poverty): only 6% of candidates achieved full marks; the examiner report notes "a few students confused relative poverty with absolute poverty" — a real, citable full-marks rate for how badly this specific reversal-prone definition is answered. (Contrast: October 2024 Q7(b), absolute poverty, was answered well by comparison.)
• June 2024 examiner report, balance-of-trade deficit vs surplus: "many simply reversed the definitions and did not get any marks."
• Fixed vs floating exchange-rate vocabulary — devaluation/revaluation belong to fixed regimes, depreciation/appreciation to floating ones — flagged as a recurring MCQ-distractor confusion in at least four series (October 2020, January 2021, January 2022).

Curriculum pointer: the relevant lesson's trap-taxonomy entry for each pair (poverty-and-inequality.ts for absolute/relative; balance-of-payments-...ts for fixed/floating vocabulary).

#EP3 — Causes ↔ effects/policies substitution

Spec ref: 4.3.1 (globalisation causes vs effects); 4.3.2.2 (patterns of trade); 4.3.4.2 (policies to reduce inequality vs causes of inequality).

Mechanism: the question asks for one category (causes, or policies) and the answer supplies a well-developed paragraph from the adjacent category instead — which earns zero, however well-written.

Evidence:

• January 2022 Q8 (factors contributing to globalisation): "not many candidates were able to entirely explain the factors identified" — many "discussed the effects of globalisation instead."
• June 2022 Q10 (government policies to reduce income inequality): "those who mentioned causes of income or wealth inequality did not attain any marks."
• January 2024 Q9 (patterns of trade): examiners record candidates "misunderstood the question and did not answer the question," blending it with the adjacent globalisation-causes content — this question scored the weakest of that series' three Section C options.

Curriculum pointer: no single trap-taxonomy fix here — this is a question-reading discipline, not a content gap. See EP9 for the related "tangential paragraph" variant.

#EP4 — Two-model / two-institution conflation

Spec ref: 4.3.6.2 (Harrod-Domar vs Lewis dual-sector model); 4.3.6.3(d)/4.3.3 (IMF vs World Bank; TNCs wrongly treated as an international institution); 4.3.5.4 (fiscal vs monetary policy); 4.3.3.1 (national debt vs current-account deficit).

Mechanism: two named things that sound structurally similar (two growth models, two institutions, two policy types, two balance-sheet concepts) get swapped for each other, and the mark scheme credits only the correct name.

Evidence:

• January 2021 Q2 MCQ ↔ January 2022 Q2 MCQ: the same two models are confused in both directions across different series — January 2021's report says candidates confused the correct answer (Harrod-Domar) for Lewis; January 2022's report says the reverse, confusing Lewis for Harrod-Domar. A genuine two-way trap, not a one-off.
• June 2023 Q9 (role of international institutions): "some candidates were quite confused about the roles of the IMF and the World Bank" and reversed them; the same report notes TNCs are sometimes wrongly listed as international institutions — "TNCs are not international institutions" and earned no credit for that content.
• October 2023 Q7(d): candidates "confused role of the IMF with the World Bank" and accessed no marks on that part.
• October 2022 Q7(d): candidates "confused fiscal policy with monetary policy" (central bank vs government) and accessed no marks.
• June 2022 Q7(b): "some candidates confused national debt with current account deficit" and accessed no marks.

Curriculum pointer: each lesson's own definitions section covers the correct pairing; codex-definitions-and-application-bank.md is the fastest place to drill the exact wording of each named model/institution.

#EP5 — Country-gate / growth-vs-development omission

Spec ref: cross-cutting Section C mechanic, evidenced most often on 4.3.4 and 4.3.6.3 essays.

Mechanism: nearly every Section C essay that asks about "a country of your choice," "a developed economy," or "a developing country" carries an explicit mark-scheme instruction capping a country-free answer at Level 3 (9/20) — verified in at least 12 of the 13 mark schemes read for questions worded that way. (Not universal: October 2023's Q8, worded generically as "a country's terms of trade" rather than "a country of your choice," carries no such gate.) On growth/development essays specifically, a second, independent gate can stack on top: January 2024's report on Q10 notes candidates were "confused between economic growth and economic development" — arguing growth when development was asked for risks the same Level-3 cap a second time, on the same essay.

Evidence: wording verified directly, January 2024 mark scheme, e.g. Q8 (income inequality essay): capped at Level 3 unless the answer "refer[s] to a developed country." Same mechanic confirmed word-for-word in October 2020, January 2021, June 2021, October 2021, January 2022, June 2022, October 2022, January 2023, June 2023, June 2024, October 2024, January 2025 mark schemes.

Curriculum pointer: name your chosen country in your opening sentence, every Section C essay, before writing anything else — this is the single cheapest fix on the whole paper for a mistake that caps 11 marks.

#EP6 — Market-orientated vs interventionist strategy misclassification

Spec ref: 4.3.6.3(a) market-orientated strategies vs 4.3.6.3(b) interventionist strategies.

Mechanism: a question asks specifically for one strategy type (usually market-orientated), and the answer lists a strategy from the other category — most often infrastructure spending, which is interventionist, not market-orientated.

Evidence:

• June 2024 Q7(e) (Indonesia, market-orientated strategies): "ensure there is a clear understanding of the difference between market-orientated and interventionist strategies" — candidates who wrote about roads and airports (infrastructure) "attained no marks" for that paragraph.
• January 2024's equivalent question: candidates "who mentioned interventionist strategies did not attain any marks" — the same trap confirmed in a second series.

Curriculum pointer: growth-and-development.ts's strategy-classification content; before writing, write "MO" or "INT" next to every strategy you plan to use and check it matches the question's own wording.

#EP7 — Diagram/data-reading direction errors

Spec ref: 4.3.2.5 (tariff diagrams); 4.3.4.2 (Lorenz curve direction); 4.3.2.1 (opportunity-cost/comparative-advantage tables).

Mechanism: the economics is understood, but the answer reads the diagram or table in the wrong direction — the wrong area, the wrong shift direction, or the wrong country's advantage.

Evidence:

• October 2020 Q5 MCQ (tariff diagram): the weakest of that series' MCQs — candidates "unable to correctly deduce the area of tax revenue from the graph."
• October 2021 Q4 MCQ (tariff diagram): candidates unable to correctly deduce that "the consumer surplus area had decreased."
• January 2023 Q5 MCQ (tariff diagram): candidates "unable to accurately understand all the different areas."
• January 2025 Q6 MCQ (steel tariff): named the explicitly weakest MCQ of that series.
• October 2023 Q4 MCQ (Georgia's Gini coefficient): candidates could not "correctly deduce from the chart" which direction the Lorenz curve shifts when the Gini coefficient falls.
• October 2024 Q4 MCQ (comparative-advantage table): named the weakest-answered question of that series — candidates "unable to correctly deduce from the data" which country held the lower opportunity cost in which good.
• January 2022 Q5 MCQ (PPF/opportunity-cost table): the boundary case where neither country benefits from specialisation was the one candidates performed worse on.

Curriculum pointer: each lesson's own diagram-reading drill; this is a "trace the axis and the shaded region with your finger before answering" discipline, not a content gap.

#EP8 — Definitional half-answer / static definition for an asked-for change

Spec ref: 4.3.2.4(b) (customs union); 4.3.5.3 (fiscal deficit reduction).

Mechanism: a definition has two required parts, or the question asks about a change in a concept, and the answer gives only the static, single-part definition.

Evidence:

• January 2022 Q7(c) (customs union): the mark scheme credits "free trade between member countries" and "a common external tariff" as two separate parts; the examiner report notes many candidates "only mentioned free trade between member countries" and accessed just one of the two marks.
• October 2020 Q7(c) (reduction in fiscal deficit): "a common response was to define fiscal deficit" rather than explain what a reduction in it means — defining the static term instead of analysing the asked-for change.

Curriculum pointer: terms-of-trade-and-trading-blocs.ts (customs union's two-part definition) and the-role-of-the-state.ts (deficit vs debt, and stock-vs-flow/change language).

#EP9 — Off-target / tangential paragraphs inside an otherwise relevant answer

Spec ref: 4.3.1 (globalisation costs); 4.3.4.2 (income inequality causes).

Mechanism: distinct from EP3 (answering the wrong category entirely) — this is a paragraph that starts on-topic but drifts into content the question didn't ask for, inside an answer that is otherwise on the right track.

Evidence:

• October 2022 Q9 (globalisation costs): "there were some who went tangential" and discussed reasons for globalisation rather than its costs.
• January 2024 Q8 (income inequality causes): examiners note candidates "offered some points that often went tangential and did not answer the question," alongside otherwise-relevant content.

Curriculum pointer: none specific — a planning discipline (one sentence per paragraph stating which part of the question it answers, checked before writing the paragraph).

#EP10 — Evaluation stalls at a two-stage chain, generic and unchained

Spec ref: cross-cutting AO4. Structurally the highest-stakes category on this paper: AO4 (evaluation) is rescaled to 30.0% of Unit 4 alone (spec p. 51 AO table, 7.5% of the full IAL ÷ 25%) — the largest single assessment objective in this unit, larger than AO1/AO2/AO3.

Mechanism: knowledge, application and analysis reach a reasonable standard, but the Evaluation band itself stops at a short, generic chain of reasoning rather than a developed judgement — the failure mode is in the evaluation specifically, not the KAA content underneath it.

Evidence: January 2024 Q8 (income inequality essay) examiner report, re-extracted directly against the primary PDF: most candidates' answers "only carried a two-stage chain of reasoning" and did not reach Level 3 KAA as a result; separately, "evaluative comments were not well written" and much of the Evaluation-band content was generic and non-chained, described as not achieving "more than Level 1."

Curriculum pointer: each curriculum lesson's own conditional-judgement drill (e.g. "depends on the size of X," "significance changes over time," "combination of factors") — the discipline is turning the generic evaluative sentence into one that names the specific condition and links it back to the specific KAA point it qualifies.

#A known weak spot without a documented mechanism — not turned into a category

The terms-of-trade Section A MCQ is named the single weakest-answered Section A question in at least 5 of 13 examiner reports checked (June 2022, October 2022, January 2023, June 2023, June 2024) — the single best-evidenced weak spot in the whole archive. The source material does not document what specifically candidates get wrong on that MCQ beyond "least well answered" (as distinct from the Q7(a) percentage-point error, which does have a documented mechanism and is EP1 above). Rather than invent a mechanism, this is flagged here as a watch item: if your own log in Part 2 shows a terms-of-trade MCQ error, describe what you actually did wrong in the "what I wrote" column — over several logged instances, your own log may document the mechanism this archive doesn't.

---

#Part 2 — Your error log

Fill in one row every time you review a timed Section A/B/C attempt (a full past paper or a single question under time pressure). Use the EP code(s) from Part 1 where one fits; write "none" and describe it in full if it doesn't. Copy the blank-row block for a new page once a sheet is full — the value of this log is in the codes repeating across many rows, not in any one row.

Worked example (delete before using):

Your log:

---

#Part 3 — Tally sheet

After every 3–4 papers logged in Part 2, transfer the count of each EP code here. Add a tally mark under the relevant column for every row in Part 2 that used that code.

---

#Part 4 — When a code keeps recurring

Not a general essay-writing lecture — just the direct next action tied to the code, once your own tally in Part 3 shows it happening more than once or twice.

• EP1 recurring: before your next timed attempt, do nothing else first — re-read terms-of-trade-and-trading-blocs.ts's percentage-point trap entry, then re-time only Section B Q7(a) on three different past papers, checking the unit of every answer before moving on.
• EP2 recurring: write both halves of the pair side by side from memory (e.g. absolute poverty / relative poverty) before you ever see a question on it — if you can't state both correctly cold, the exam won't go better.
• EP3 or EP9 recurring: before writing a single paragraph in Section C, underline the exact noun the question asks for (causes? effects? policies?) and write it at the top of your plan; check every paragraph against that noun before writing it, not after.
• EP4 recurring: drill the two confused items as an explicit contrast pair (Harrod-Domar vs Lewis; IMF vs World Bank; fiscal vs monetary) rather than revising them separately — the archive's evidence is that they're confused because they're revised in isolation.
• EP5 recurring: make naming the country literally the first thing you write in every Section C essay, before the introduction sentence — treat it as non-negotiable, not stylistic.
• EP6 recurring: before Q7(e) or a growth/development essay, list every strategy you plan to use and label each "MO" or "INT" against the spec's own two categories before writing.
• EP7 recurring: for tariff-diagram and Lorenz-curve MCQs specifically, physically trace the relevant axis and shaded region before selecting an answer — don't answer from memory of what the diagram "usually" shows.
• EP8 recurring: for any "explain the term" part, check whether the definition has two required parts (most do) and whether the question asks for a static concept or a change in it — answer the one actually asked for.
• EP10 recurring: this is the highest-value code to fix, given AO4's 30% weighting in this unit — for your next three essays, write the Evaluation band as full sentences of the form "this matters more/less because [named condition], which affects [named KAA point]," and check afterwards that you did not write a sentence starting "it depends" without finishing the thought.

---

#Sources, coverage and honest gaps

Every category above is drawn from research/veridian/WEC14-verified-facts.md, itself checked against 14 real WEC14/01 exam series (October 2020–January 2025) and 13 matching examiner reports, plus one narrower June 2025 check (see that file's "Series coverage log" for the exact scope and its own flagged extraction issues on the June 2022, June 2023 and June 2024 examiner-report PDFs). This tracker does not cover:

• Any series without a matching examiner report — June 2021 has a question paper and mark scheme in the archive but no examiner report, so no candidate-performance pattern from that series appears above.
• 4.3.2.4(d) (conflicts between trading blocs and the WTO) — the source document itself flags this spec point as not yet matched to a specific examiner-reported error.
• Any mistake pattern not directly evidenced by a named series and question in the source material. If your own Part 2 log surfaces a repeating mistake that isn't one of the ten codes above, it is real data your own practice has generated — log it as "Other" in Part 3 rather than forcing it into the nearest existing code.