Diagram Masterclass

Provenance (written 2026-09-29). This pass checked, before writing anything, whether WBS13 (Business, Unit 3) genuinely credits diagrams/graphs as their own distinct mark-scheme skill, per this session's brief. Sources read in full first: research/veridian/WBS13-verified-facts.md, the existing content/veridian/curriculum/wbs13/decision-trees-cpa-and-contribution.ts and forecasting-and-investment-appraisal.ts lessons plus their codex-deep-dive-*.md companions, and every file already in content/veridian/technique-hub/wbs13/ — most relevantly forge-tariff-4-mark-questions.md, which already contains a full, independently-verified deep-dive on the Construct command word. Rather than trusting that document's own summary, this pass independently re-ran the same kind of primary-source check it describes: grep -n -i "QS3" and related quantitative-skills-code searches across /Users/neosteinhoff/Documents/SOL/Papers/IAL_Paper3_Complete/Master_Consolidated_Docs/ Business_Paper3_Mark_Schemes.md and Business_Paper3_Past_Papers.md (a pdftotext-derived transcription of every real Pearson WBS13 mark scheme/question paper from January 2020 to October 2024 — the same archive WBS13-verified-facts.md itself already cites and treats as genuine primary-source material, e.g. its Summer-2024/SWOT and Burger-King/CPA entries). Every quotation and QS code below was read directly from that file this pass, not copied from another document's transcription of it. The verdict below is a real "yes, but narrowly" — not the simple yes/no the brief anticipated — so the document is structured around the evidence that produces that nuance, not around a template.

#WBS13 Diagram/Graph Masterclass — the real map, not a repeat

#Verdict, stated plainly

Diagrams and graphs genuinely are a separately-markable skill on this paper — but only for ONE diagram type, and that type is already fully covered elsewhere in this hub. Every other spec-listed "diagram" on WBS13 (decision trees, critical path networks, portfolio-analysis matrices) is tested as a calculation-and-interpretation skill embedded inside a 20-mark extended-writing question, not as a stand-alone drawing task with its own accuracy marks. One more piece of spec content that sounds diagram-shaped — scatter graphs and the line of best fit — has never once appeared as a real WBS13 question in any series checked. This document's job is therefore not to re-teach diagram mechanics the hub already covers well, but to give you the map: which of the paper's diagram-shaped content is actually worth drilling as a drawing skill, and which just looks like one.

#The evidence trail: quantitative-skills codes, not vibes

Every WBS13 mark scheme states, for each extended-writing question, exactly which of the spec's nine Quantitative Skills (Appendix 7, QS1–QS9) it is assessing. One of those nine is explicitly the diagram/graph skill:

QS3: Construct and interpret a range of standard graphical forms.

This is the single most useful piece of evidence for this task, because it turns "does this feel like a diagram question" into something checkable: does the real mark scheme actually tag this question QS3, or not. Searching every occurrence of QS3 across the full Jan-2020–Oct-2024 archive returns exactly five hits, across three different question types:

That's the whole QS3 list for the entire archive. Nothing else on the paper is coded QS3 — not decision trees, not portfolio analysis, not Ansoff's Matrix — confirmed directly, not assumed:

Read together, this is a precise, primary-source-grounded answer to the brief's actual question, not an assumption: QS3's "construct" half has only ever been invoked for the supply-and-demand diagram; its "interpret" half has only ever been invoked for a critical-path network that the extract hands you, complete, to read; and every other diagram-shaped spec point on this paper is tested purely as QS8/QS9 information-handling — no graphical code attaches to it whatsoever.

#Row 1 — Supply and demand (Construct, 4 marks): don't repeat, go read the existing deep-dive

This is the paper's only stand-alone diagram-drawing mark, and it is already treated at masterclass depth in forge-tariff-4-mark-questions.md, Part 2 — three independently re-verified real questions, the Knowledge/Application/Analysis mark breakdown, and every examiner-confirmed error pattern (axis labels graded literally as "Price"/"Quantity" only, the original equilibrium having to be explicitly labelled and not just visually present, the new equilibrium having to be redrawn and relabelled rather than left at the intersection point, and the advice to use a ruler). Go there for the mechanics; nothing below repeats them.

#Row 2 — Critical path networks: interpreted, not drawn

Both real CPA questions found (Coca-Cola, June 2022 Q2; Burger King, Summer 2024 Q3) hand you the network's own activity list and durations inside the extract — Extracts F/G and G/H respectively — as a diagram image, not as something you construct from a blank page. What QS3 credits here is reading that given diagram correctly enough to calculate the right project duration and float, then using those numbers in an evaluative argument. The cleanest proof that the drawing itself carries no separate mark: the Burger King examiner report describes a real 18/20 script that correctly used "a network diagram" and got the calculation right, and lost marks purely on an underdeveloped conclusion — not on anything about the diagram's construction or presentation. The labelling conventions that matter for this skill (nodes as numbered circles, activities as labelled arrows, EST built forward by taking the later of converging routes, LFT built backward by taking the earlier of diverging routes, float = LFT − EST − duration, critical path = the zero-float chain) are already taught from first principles, with the Coca-Cola and Burger King citations attached, in decision-trees-cpa-and-contribution.ts and its codex-deep-dive companion — cross- referenced here, not re-derived.

What this means for revision time: don't drill "drawing an accurate CPA network" as its own exam skill the way you would S+D — no mark scheme has ever rewarded that specifically. Drill reading a given network's activity list correctly, doing the EST/LFT/float arithmetic without error, and then writing an evaluation that engages with this business's numbers (32 weeks for Coca-Cola, 11 weeks and Activity C's 1-week float for Burger King) rather than a generic "CPA is useful because…" answer.

#Row 3 — Decision trees: not even coded as graphical

The spec's own wording for 3.3.3.3a — "construct/interpret" — reads as if it should sit alongside S+D and CPA under QS3. The one real decision-tree question in the archive, Center Parcs (Oct 2024 Q2), doesn't: its QS line reads "QS1, QS8, QS9", meaning the mark scheme itself treats this purely as a calculation (QS1: the EMV arithmetic) plus an information-interpretation skill (QS8/QS9), never touching the graphical-forms code at all. The mark scheme's own indicative content even includes a prose point crediting the observation that "a decision tree… is a visual tool that may help… management… discuss its strategy" — but that's a written argument point about DTA's nature, markable like any other Knowledge bullet, not a mark for drawing the tree accurately. The square (decision node) / circle (chance node) notation this course teaches for decision trees is standard, universal DTA convention — not something reproduced from a WBS13 mark scheme's own diagram, since real WBS13 diagrams are supplied as image extracts, not machine-readable text (the same caveat this hub's CPA content already carries).

What this means for revision time: the EV-then-net-gain arithmetic is where every real mark lives on this sub-topic (confirmed: QS1). Time spent perfecting hand-drawn tree aesthetics is time not spent on the thing the mark scheme actually rewards.

#Row 4 — Portfolio analysis and Ansoff's Matrix: pure prose classification

Both checked directly. Pets at Home's portfolio-analysis question (Oct 2022 Q1d) is coded QS8, QS9 — no QS3. Five Guys' Ansoff's-Matrix question (Oct 2023 Q1e, "assess the possible benefits to Five Guys from launching its new clothing merchandise") is coded QS8, QS9 — no QS3 either. Neither matrix is ever drawn by the candidate for marks; both are classification-and- justification frameworks applied in prose to the extract's real figures (Pets at Home's 45% market share and 15% revenue growth marking pet accessories as a cash cow; Five Guys' existing customer base marking clothing merchandise as market development on Ansoff's grid). This is genuinely different from how these same two matrices sometimes get taught as "draw the grid, plot the product" — that presentation may help you reason about the classification, but it earns nothing extra on this paper that the classification itself, stated in words, doesn't already earn.

#Row 5 — Scatter graphs and the line of best fit: spec content, zero real appearances

3.3.3.1b is explicit, numbered spec content, and this course's own forecasting lesson already teaches it honestly, including the fact that the line is "drawn by eye… no formula is required" on this paper. Searching for scatter across the ENTIRE archived question-paper and mark-scheme text (Jan 2020–Oct 2024) returns zero results — not "not yet found," but genuinely absent from every question and every mark scheme in the archive. The two real questions this course has verified for this sub-topic (Spotify, Jan 2025 Q1c; Pets at Home, Oct 2022 Q1c) both test the limitations of quantitative forecasting as an 8-mark Discuss, using words, not a plotted graph. This is not the same claim as "scatter graphs will never appear" — a spec point can sit unexamined for many series and still be fair game — but it is the honest, checked state of the evidence: no primary-source WBS13 question has ever asked a candidate to plot a scatter graph or draw a line of best fit for marks.

#Honest gaps in this specific check

• Porter's Strategic Matrix and Porter's five forces were not individually re-checked for a QS3 tag this pass, unlike Ansoff's and the Boston Matrix (both of which were). Given both are the same genre of prose-classification framework applied to extract data (confirmed elsewhere in WBS13-verified-facts.md for the Superdry/Porter's-Strategic-Matrix and Apple-Watch/five-forces questions), the same QS8/QS9-only pattern is the reasonable expectation — but that is an inference from a consistent pattern across every other framework checked, not an independently re-verified fact for these two specifically, and it is flagged as such rather than stated as confirmed.
• Jan 2025 and Oct 2025 series sit outside the Jan-2020–Oct-2024 consolidated archive this pass searched by full text. WBS13-verified-facts.md's own topic-by-topic mining of both series (Spotify/gearing, Tesla/labour-productivity, Superdry/Porter's Strategic Matrix, Mears Group/consultation) found no Construct or scatter-graph question in either — but that mining wasn't itself a dedicated QS-code text search of those two PDFs the way this pass ran across the older archive, so a future citation-currency pass could usefully re-run the same QS3/scatter grep directly against those two documents once downloaded, rather than relying on topic-by-topic coverage alone.
• This document does not reproduce the S+D Construct mechanics, CPA network-diagram mechanics, or decision-tree notation in full — each is already taught properly elsewhere, and the point of this pass was the map between them, not a fourth copy of any one of them.