Technique Masterclasses

Statistics 1

Spaced Repetition Schedule

Provenance: compiled 2026-09-29. Built primarily from signal-frequency-and-boundaries.md §1 of this same folder (the topic-frequency table this schedule's tiering is derived from, row by row) and research/veridian/WST01-verified-facts.md §5 (per-topic verified content, read in full for every named trap cited below), §7 ("topics with enough verified material for a genuine standalone lesson"), §6 (the cross-cutting "show that" pattern), and §2b (the AO3 weighting that funds spec 1.1). Cross-checked against forge-battle-card.md §4–§5 (the top-5 danger-warning citations, reused here only where they add a repeatability count the facts bank's own §5 doesn't already give) and forge-exam-technique-guide.md (timing, mark-coding, self-mark checklist), so neither is repeated here. The 13 curriculum lesson files in content/veridian/curriculum/wst01/ were checked by filename against this folder's topic-brief list to confirm the tier mapping in §2, not re-read in full — their own worked exemplars are not reproduced below. content/veridian/question-bank/wst01.ts was checked and holds two items (a June 2019 Q2(c) box-plot/outlier correction, a June 2019 Q5(e) expectation-vs-win-probability item); neither is duplicated below. Every repetition count and day-interval in §3 is this course's own scheduling construction, reusing the Day-0/2/5/10/18/30 six-cycle widening-interval convention already established for this course in content/veridian/technique-hub/wbs11/forge-spaced-repetition-schedule.md §3 and reused for WMA11 and WME01 — flagged as a reused calendar convention, not a new finding. The tier logic itself (§1–§2) is not copied from those two documents: WST01's own examiner-report archive is 6 series deep, not 17, so a repeatability bar built for a 17-series paper does not transfer directly — see §1 for how this schedule adapts it. No new PDFs, web searches, or primary-source verification were performed for this document. Not affiliated with or endorsed by Pearson Edexcel. No likelihood percentage or grade boundary appears anywhere below — signal-frequency-and-boundaries.md §2 confirms none exists in this course's source record for this paper.

WST01 — Spaced Repetition Schedule (Unit S1: Statistics 1)

0. What this is, and what it deliberately doesn't repeat

This document answers one question: given WST01's real, uneven distribution of exam coverage across its six spec sections, how many times should a student revisit each one before the exam, and in what order? Nothing below re-teaches content. It does not reproduce:

What it adds: an actual repetition count per spec section, built by reading three real signals from this paper's own verified record — each section's confirmed sighting count, its recovered mark tariff where one exists, and how concentrated its named error is (whether the same mistake recurs almost every time the section is tested, or only some of the time) — rather than assuming every section deserves the same number of passes. No other WST01 technique-hub document does this.

1. Why repetition count should track three signals, not raw frequency alone

signal-frequency-and-boundaries.md §1 gives one finding cleanly: across the 6 series with a fully-read examiner report, Section 2 (data representation and summary) is confirmed tested in 6 of 6 — the only section confirmed on literally every series checked. A schedule built from that number alone would rank the other five sections purely by their own "of 6" count and stop there. That would miss something real in this paper's own record.

First, a repeatability bar built for a larger archive doesn't transfer to this paper's smaller one without adjustment. This course's WME01 schedule uses "a single named error confirmed across 3 or more series" as its strongest signal, against a 17-series archive — 3 of 17 is a real, rare finding there. WST01's fully-read examiner-report archive is only 6 series deep (signal-frequency-and-boundaries.md §1 methodology note). Against a base of 6, "confirmed in 3 series" is not a rare signal — it is half the archive. Applying the 17-series bar unchanged would misread ordinary WST01 sections as exceptional. This schedule instead asks a sharper question per section: of the series where the section was itself confirmed tested, what fraction also show the same specific error? Call this concentration. A trap confirmed in 3 of 6 series overall is a different finding depending on whether the section itself was only tested 4 times (3 of 4 = a near- guarantee whenever it appears) or tested all 6 times (3 of 6 = roughly half).

Second, doing this arithmetic surfaces a genuine, checkable finding: two sections' headline traps are confirmed in every single series where the section itself was confirmed tested — not "often," but every time. Section 3's numerator/denominator conditional-probability confusion is cited in Jan 2021 Q1(c), Oct 2021 Q4(c), Jun 2022 Q4(b) and Jan 2023 Q2(d) [WST01-verified-facts.md §5, Section 3] — the same four series signal-frequency-and-boundaries.md §1 names as Section 3's own "4 of 6" confirmed-tested list. Section 4's context-interpretation weak point is cited in Oct 2021 Q2(d), Jun 2022 Q2(d), Jan 2023 Q6(a) and Jun 2024 Q4(d) [WST01-verified-facts.md §5, Section 4] — again exactly the four series that section's own "4 of 6" count names. Both traps sit at 4-of-4 concentration: whenever the fully-read record shows the section examined at all, this exact error is independently flagged. That is a stronger, more specific finding than either section's raw "4 of 6" alone, and it is why both outrank Section 5 below despite Section 5's higher raw frequency.

Third, where this course's own sources disagree with each other on a citation count, this schedule states the disagreement rather than picking a number. forge-battle-card.md §4 states the discrete-random-variable squaring error — forgetting to square aa in Var(aX+b)=a2Var(X)\text{Var}(aX+b)=a^2\text{Var}(X) — as "seen in 3 of 6... series," naming Jan 2021 Q4, Jun 2024 Q2(c) and Jun 2024 Q3(d). Those three citations name only two distinct series (Jan 2021 and Jun 2024 — the error recurs twice within the one Jun 2024 paper, at Q2(c) and Q3(d)). Read as distinct-series concentration against Section 5's own 5-of-6 topic frequency, that is 2 of 5, not 3 of 6. Both counts are real (three confirmed instances, two confirmed series) and this schedule uses the more conservative one — 2 distinct series — for tiering, flagging the battle card's phrasing as the looser of the two rather than silently adopting whichever number ranks the topic higher.

None of this licenses skipping a thin section. Every spec section is compulsory, and Section 1 (mathematical models) carries real marks — 15–20 of 75, via AO3, spread through the rest of the paper [WST01-verified-facts.md §2b] — despite anchoring zero standalone questions in the archive reviewed. What the three signals above license is an honest allocation of repeated practice: a section with a trap confirmed every time it's tested has a genuine, narrow pattern worth drilling again and again; a section with real marks but no standalone question and no repeatable trap of its own has nothing of that shape to rehearse beyond one accurate pass.

2. The four tiers, built from each section's own real evidence

Tier 1 — the two sections that each win outright on a different real signal

Tier 2 — a headline trap confirmed every single time the section was itself confirmed tested

Tier 3 — real frequency comparable to Tier 1, but the headline trap's own citation count is thinner than it first looks

Tier 4 — confirmed content, but no standalone question anywhere in the archive to drill against

3. The six-cycle schedule

This reuses the Day 0/2/5/10/18/30 six-cycle widening-interval calendar already established for this course in wbs11/forge-spaced-repetition-schedule.md §3 and reused for WMA11 and WME01 — a scheduling convention, stated here as one, not a new finding. If fewer than 30 days remain before the exam, compress by dropping the latest cycles first and keep Cycle 1 and Cycle 2 intact, since Tier 4's one pass and Tier 1's first pass both live there.

CycleDaySection 2 (Tier 1)Section 6 (Tier 1)Section 3 (Tier 2)Section 4 (Tier 2)Section 5 (Tier 3)Section 1 (Tier 4)
10Pass 1Pass 1Pass 1—Pass 1Only pass
22Pass 2Pass 2—Pass 1——
35Pass 3Pass 3Pass 2———
410Pass 4Pass 4—Pass 2Pass 2—
518Pass 5Pass 5Pass 3———
630Pass 6Pass 6—Pass 3——

Reading this as repetition counts rather than a calendar: Section 2 and Section 6 each get six passes — every cycle — matching Tier 1's two independent wins (highest raw frequency; heaviest tariff plus densest trap taxonomy). Section 3 and Section 4 each get three passes, spaced at Days 0/5/18 and 2/10/30 respectively, reflecting Tier 2's 4-of-4 trap concentration even though their raw frequency (4 of 6) sits below Section 5's. Section 5 gets two passes (Days 0 and 10), reflecting its real 5-of-6 frequency without overstating its trap's own thinner, two-series citation. Section 1 gets exactly one pass, in Cycle 1 only, consistent with §1's reasoning: a section with zero standalone questions in the archive has no repeatable exam-question shape to rehearse beyond one accurate encoding pass.

4. Judgement calls, stated explicitly

Section 3's two thinner traps (independence-vs-mutually-exclusive; the without-replacement denominator error) are not given their own slots. Neither individually reaches the 4-of-4 concentration that earned Section 3 its Tier 2 place, and both are the same underlying skill area (probability mechanics) as the headline conditional-probability trap. Every Section 3 pass above (Days 0, 5, 18) should therefore end with two specific checks alongside the main conditional- probability drill: is independence being assumed rather than tested (the ER's own "the usual confusion" phrase, Oct 2021 Q1(b)), and, on any without-replacement question, does the denominator actually shrink on every draw rather than staying fixed (Jun 2022 Q3(d)).

Section 4's three thinner items (extrapolation-reliability; the no-fraction rule; PMCC accuracy and its invariance under coding) are folded into Section 4's own three passes the same way. Every Section 4 pass (Days 2, 10, 30) should end with: naming the specific value and range boundary when judging extrapolation reliability, not a bare "unreliable" (Jun 2022 Q2(e), Jan 2023 Q6(c)); writing the final regression line as a decimal, never a fraction (Jun 2022 Q2(c)); and, on any PMCC "show that," carrying one extra decimal place through the working (Jun 2024 Q4(c)) while remembering that rr itself is unaffected by linear coding (Jan 2024 mark scheme Q2(d), a mark-scheme-only citation, not from a fully-read examiner report).

The "show that" answer-discipline habit is not given its own slot anywhere in §3, because the evidence itself says it isn't one section's property. Pearson's own general comments name it as paper-wide — "A general comment, which applies to other questions as well, is that many students do not really understand that evidence has to be given for 'show that' questions" (Oct 2021, general) — and the facts bank corroborates this with named instances inside Section 2 (Oct 2021 Q3(b), IQR), Section 4 (Jun 2024 Q3(c) and Q4(c)), Section 5 (Jan 2023 Q3(a) and Q4(i)(a)) and Section 6 (Oct 2021 Q6(a)) [WST01-verified-facts.md §6]. This is broader evidence than any one topic's own trap, so unlike a habit tied to a single context, it is folded into every pass above, regardless of section: each pass ends with the same question — is the intermediate step actually written down, not just the target value repeated back.

Section 1's single pass is not its only real exposure. Because 1.1's marks are funded through AO3 rather than a standalone question, the modelling-assumption skill it covers is also exercised inside any Tier 1 or Tier 2 pass that reaches a "conclusion in context" or assumption-justification part — the same four trigger points forge-battle-card.md §3 already names. Cycle 1's dedicated pass is where the concept itself (what a model is, why "the die is fair" is an assumption and not a fact) gets taught explicitly, since the archive gives no standalone worked example to anchor it to [WST01-verified-facts.md §5, Section 1]; the other five cycles rehearse it indirectly, not by neglect.

The two source disagreements flagged in §1 (the "3 of 6" vs 2-distinct-series discrete-RV count; the "5 examiner reports" phrasing in WST01-verified-facts.md §5's Section 3 and Section 4 prose against signal-frequency-and-boundaries.md's own "6 of 6 fully-read" archive total) are stated there, not resolved by rounding either number up or down. The second disagreement is most likely explained by timing — the "5 examiner reports reviewed" phrasing in §5's per-topic write-ups predates the 2026-09-25 spot-check addendum that added a sixth (June 2019) fully-read report [WST01-verified-facts.md §8] — but this schedule does not assert that explanation as confirmed; it uses each source's own citations exactly as given.

5. Honesty notes