# Quizzes > Generate scored practice quizzes from your own material, with an explanation for every wrong answer. Use when authoring multiple-choice questions, plausible distractors, or trivia. - Skill: `poorvith-mp/quizzes` (Agent Skill, multi-file: 3 files) - Install (CLI): `npx skillmds@latest add poorvith-mp/quizzes` - Raw SKILL.md: https://api.skillmd.com/api/skills/poorvith-mp/quizzes/raw - Safety review: pending - Works with: Claude Code, Claude.ai, OpenAI Codex - Category: Coding & Dev Tools - Author: poorvith-mp (https://skillmd.com/u/poorvith-mp) - Updated: 2026-09-17 - Page: https://skillmd.com/skills/poorvith-mp/quizzes --- # quizzes ## Core Philosophy A practice quiz is not an administrative grading hurdle; it is a **diagnostic instrument for targeted active recall**. Poorly written quizzes feature trivial recall stems, laughable distractors that give away the answer by elimination, or ambiguous questions where two options could be right. High-yield diagnostic quizzes are precision-engineered: every question isolates a specific conceptual distinction, every incorrect distractor embodies a documented cognitive misconception, and every question provides comprehensive rationales explaining not just why the right answer is correct, but *why each specific wrong answer was chosen and why it fails*. --- ## 4-Step Diagnostic Quiz Formulation ### Step 1: Isolating the Diagnostic Concept (The Stem) 1. **The Affirmative Stem Rule**: - Write clear, focused scenario-based stems that state the exact problem without trickery: - *Bad*: "Which is true about database indexes?" (Too broad; invites guessing). - *Good*: "A PostgreSQL query executing `SELECT * FROM orders WHERE customer_id = 42 ORDER BY created_at DESC LIMIT 10` performs a slow sequential scan despite an existing B-Tree index on `customer_id`. Which index definition eliminates the sorting bottleneck?" 2. **Context Sufficiency**: - The test-taker should be able to anticipate the answer *before* reading the 4 options. ### Step 2: Engineering Plausible Distractors (The Misconception Map) 1. **The 3 Distractor Archetypes**: - **Distractor A (The Partial Truth)**: Correct concept applied to the wrong layer or scope. - **Distractor B (The Inverse / Opposite)**: Common student reversal of cause and effect. - **Distractor C (The Lookalike / Jargon Trap)**: Uses familiar technical buzzwords in an incoherent configuration. 2. **Grammatical & Length Parity**: - Ensure all 4 options share roughly identical length, tone, and grammatical structure so the correct option does not stand out visually. ### Step 3: Comprehensive Distractor Rationales 1. **The 4-Way Explanation Standard**: - For every question, write an exhaustive debrief: - **Option A (Correct)**: Explains the exact governing mechanism. - **Option B (Incorrect)**: Explains why this option is wrong and diagnoses what misconception led the student to pick it. - **Option C (Incorrect)**: Explains the underlying boundary condition failure. - **Option D (Incorrect)**: Explains the contextual mismatch. ### Step 4: Scored Practice & Mastery Thresholds 1. **The 80% Gate**: - A concept is not cleared for production or exam readiness until scored at $\ge 80\%$ on an unprompted diagnostic quiz. 2. **Re-Testing Lapsed Concepts**: - Missed questions must be re-queued with variant numerical parameters or scenario framing 48 hours later. --- ## Deliverable Format: Production Diagnostic Quiz Suite ```markdown # Diagnostic Quiz: Database Concurrency & Isolation Levels ### Question 1: In PostgreSQL (Read Committed isolation level), Transaction A reads a row with balance = $100. Concurrently, Transaction B updates that same row to balance = $150 and commits. If Transaction A re-reads the exact same row within its active transaction block, what value will it observe, and what concurrency anomaly is demonstrated? - **A)** $100; Non-repeatable Read - **B)** $150; Non-repeatable Read (Correct) - **C)** $150; Dirty Read - **D)** $100; Phantom Read --- ### Detailed Answer & Distractor Deconstruction: - **Correct Option: B ($150; Non-repeatable Read)** - *Explanation*: In Read Committed isolation, each query within a transaction sees a new snapshot taken at query start. Because Transaction B committed before Transaction A's second query executed, Transaction A observes the updated value ($150). This phenomenon—reading different values for the same row within a single transaction—is the textbook definition of a Non-repeatable Read (Fuzzy Read). - **Why Option A is Incorrect ($100; Non-repeatable Read)**: - *Misconception Diagnosis*: You recognized the name of the anomaly, but assumed the transaction took a single snapshot at transaction start. Single-transaction-level snapshots occur in *Repeatable Read* and *Serializable* isolation, not *Read Committed*. - **Why Option C is Incorrect ($150; Dirty Read)**: - *Misconception Diagnosis*: A Dirty Read occurs when a transaction reads *uncommitted* mutations from a concurrent transaction that might later roll back. Because Transaction B successfully committed *before* Transaction A read the data, this is not a dirty read. PostgreSQL never permits dirty reads under any isolation level. - **Why Option D is Incorrect ($100; Phantom Read)**: - *Misconception Diagnosis*: A Phantom Read occurs when a range query (`WHERE amount > 50`) returns newly inserted or deleted rows that match the predicate. Changing an existing row's scalar value is a non-repeatable read, not a phantom read. ``` --- ## Worked Example: Calibrating a Cloud Security Certification Quiz - **Problem**: A corporate AWS security training program had a 95% pass rate on internal quizzes, but 60% of engineers failed the official AWS Certified Security exam. - **Diagnosis**: Internal quizzes used binary true/false questions and obvious distractors (e.g. "Option D: Disable all firewalls"). - **Intervention**: 1. Rewrote 50 quiz items into scenario-based architectural trade-off questions. 2. Engineered distractors around real-world IAM edge cases (e.g. SCPs overriding resource-based policies vs identity-based policies). 3. Mandated reading the comprehensive distractor breakdown for every attempt. - **Outcome**: First-time pass rate on the official AWS certification leaped from 40% to 88%. --- ## Verification Checklist - [ ] Question stems provide sufficient situational context without ambiguous phrasing. - [ ] Every question features 4 plausible, grammatically balanced options of similar length. - [ ] All 3 distractors map to documented student conceptual errors. - [ ] Exhaustive rationales explain why each wrong answer is incorrect. - [ ] Minimum 80% passing threshold enforced before clearing topics. --- ## Anti-Patterns - **The Obvious "Throwaway" Distractor**: Including a humorous or absurd option (e.g. "C: Unplug the server"), reducing a 4-option quiz to a 3-option or 2-option guess. - **Negatively Phrased Trick Questions**: Asking "Which of the following is NOT an invalid way to..." which tests semantic parsing rather than technical mastery. - **Answer Keys Without Explanations**: Providing only "Answer: B" with zero diagnostic breakdown of why B is right and why A, C, and D are wrong.