LESSON 14.2 — Concept-Family PYQ Sets
A. Standard Map
| Stage | Input | Output | Purpose |
|---|---|---|---|
| Taxonomy lock | Course chapter list (Ch 1–13) | Concept-family labels | Prevent fragmenting frequency |
| PYQ harvest | 5–7 year papers | Question rows tagged by family | Build cross-year dataset |
| Variant classify | Each question | Direct / Application / Twist tag | See trap evolution |
| Set design | Family + variant gaps | Escalating practice set (3–5 Q) | Close weakest variant first |
| Stop rule | Accuracy + time gates | Family status: Open / Maintained / Closed | Avoid infinite drilling |
B. Why It’s Used
Concept-family sets convert scattered PYQs into skill-closing drills — the bridge between L14.1 priority map and exam execution.
C. Mechanism in Words
- Define the family: Pick one examinable concept (e.g., PCU vs ECS, PERT expected time, U-value assembly) — not a chapter title alone.
- Pull all PYQs for that family across years into one list (ignore year order initially).
- Label each question D / A / T and note trap type (unit swap, definition inversion, partial truth MSQ).
- Sort weakest variant first: If Twist fails but Direct passes, the set order is Direct → Application → Twist.
- Run timed sets: Minimum 3 questions per session; record error class (concept / process / arithmetic / reading).
- Apply stop criteria: ≥80% accuracy on Twist tier within time gate → family moves to maintenance (1 Q/week).
D. Core Concept Explanations
D1. Defining Concept Families
A concept family is the smallest examinable idea that repeats across questions with different surface stories. “Transport Planning” is too broad; “PCU aggregation for mixed traffic” is a family.
D2. Grouping PYQs by Concept Instead of Year
Year-wise analysis (L14.1) ranks topics; concept-family grouping ranks skills within a topic. Three years may each test PERT with different variants — one family, three traps.
D3. Trap Evolution Across Years
Setters preserve the concept but shift trap: MCQ definition → MSQ multi-step → NAT numeric. Track trap type: wrong base, wrong unit, reversed comparison, MSQ partial truth.
D4. Designing Escalating Practice Sets
Template set (45 min): Q1 Direct → Q2 Application → Q3 Twist. Do not start with Twist — it hides whether the gap is recall or reasoning.
D5. Deciding Stop Criteria for a Concept Family
| Status | Enter when | Maintenance |
|---|---|---|
| Open | Twist accuracy <80% within time gate | 3 Q/session, 2×/week |
| Maintained | Twist ≥80% within gate | 1 Q/week |
| Closed | Maintained 3 weeks + mock proof | Monthly spot check |
E. Worked Workflows and Practice Drills
Family: PERT expected time (TE)
| Year | Pattern | Variant | Trap |
|---|---|---|---|
| 2020 | O, M, P → find TE | Direct | Mean of three instead of (O+4M+P)/6 |
| 2022 | TE given → critical path | Application | Longest TE ≠ critical path |
| 2024 | MSQ on TE + variance + CP | Twist | Float confusion |
Escalating set: (1) A: O=2, M=5, P=14 → TE=6. (2) Three-activity network → CP. (3) MSQ proof each option.
Stop: Q3 wrong on float → tag process error (L14.4), add float drill, re-run Twist only.
F. Design Criteria
| Element | Key Fact |
|---|---|
| GATE AR | 100 marks; 65 Q |
| MCQ penalty | −⅓ / −⅔ |
| MSQ/NAT | No negative marking |
Code Comparisons
| Approach | Advantage | Risk |
|---|---|---|
| Year-first | Trend visible | Misses cross-year traps |
| Concept-family | Trap evolution visible | Needs taxonomy |
G. Application Zones
Daily 45–60 min blocks; weekly tie to Tier matrix; mocks use timed sets only.
H. Common Confusions
- Family vs chapter: Chapters are containers; families are exam skills.
- Variant mis-tagging: MSQ is at least Application, not Direct.
- Year-only revision: Misses trap evolution visible cross-year.
- Set size inflation: >5 Q/family/day → fatigue without signal.
- Stopping too early: One Twist correct ≠ family closed.
- MSQ greed: Selecting unprovable third option.
I. Compare & Contrast
| Ad-hoc PYQ | Structured family set |
|---|---|
| Answers attempted | Skill state + next action |
J. Memory Hooks
Family not chapter · Direct→App→Twist · Max 6 open families
K. Revision Ladder
- Explain workflow · 2. Complete Section D drill · 3. Mini-check ≥4/5 · 4. Ten PYQs one cluster · 5. Weekly audit
L. Exam Traps
| Trap | Description | Correction |
|---|---|---|
| Chapter-level grouping | All Structures as one family | Split PERT, beams, loads |
| Random set order | Twist before Direct | Direct → Application → Twist |
| Single-year proof | “2023 Q correct” | Cross-year variant gate |
| Formula without condition | TE when question asks float | Read requested output |
| MSQ greed | Uncertain third option | Provable TRUE only |
| NAT carry-over | MCQ tricks on NAT | Unit + entry checklist |
| Time gate ignored | 15 min on 1M warm-up | Cap Direct at 5 min |
| No error tag | Repeat float mistake | L14.4 notebook |
| Family sprawl | 40 open families | Max ~6 on Tier 1 |
| Closing on mock luck | One mock Q | 3-week maintenance |
M. Answer-Writing Cues
Direct: “Family = [X]. Rule = [one line]. Given = [ ]. Asked = [ ].”
Application/NAT: “Units → substitute → compute → magnitude check → entry format.”
MSQ: “Option A: T/F + proof. … — select proven only.”
N. PYQ Integration (2007–2026 verified)
| Family | Variant progression | Drill |
|---|---|---|
| PERT/CPM | D → A → MSQ | Include float check |
| PCU/ECS | Def → numeric → MSQ | Separate families |
| U-value | Formula → assembly | Unit drill mandatory |
O. Mini-Check — Lesson 14.2 (5 Questions)
Q1 (MCQ): A concept family is:
(a) Any syllabus chapter (b) Smallest repeating examinable skill (c) One year’s type (d) Mock section
Answer: (b)
Q2 (MCQ): Correct escalating order:
(a) Twist→Direct→App (b) Direct→App→Twist (c) Random (d) MSQ only
Answer: (b)
Q3 (MCQ): Move to Maintained when:
(a) One Twist correct (b) Twist ≥80% within time gate (c) Notes read (d) 5 papers
Answer: (b)
Q4 (MSQ): Valid practices?
(a) Tag D/A/T (b) One Transport family (c) Track traps cross-year (d) Stop at easy Direct (e) Cap ~6 open families
Answer: (a), (c), (e)
Q5 (MSQ): Direct✓ App✓ Twist✗ (float). Next?
(a) Close (b) Process error + float drill (c) Skip (d) Formula sheet only
Answer: (b)