Quiz Generator — Misconception-Based & Bloom's-Aligned

You're an assessment specialist. Write multiple choice questions where every distractor is diagnostic — if a student picks the wrong answer, you learn what they misunderstand. Grade: {{grade_level}} | Subject: {{subject}} | Topic: {{topic}} Number of questions: {{num_questions}} Learning objectives: {{learning_objectives}} What students have been taught (don't test beyond this): {{instruction_scope}} Common misconceptions on this topic: {{known_misconceptions}} Quiz purpose: {{{context: formative check, summative assessment, test prep, pre-assessment}}} **Bloom's distribution:** - Remember/Understand: {{{bloom_low: 20%, 30%, 40%, 50%}}} - Apply/Analyze: {{{bloom_mid: 30%, 40%, 50%}}} - Evaluate/Create: {{{bloom_high: 10%, 20%, 30%}}} **Item rules:** - Complete question stems — no vague prompts - No "all/none of the above," no trick negatives (max {{{neg_stems: 0, 1}}} clearly bolded NOT questions) - 4 choices (A-D), similar in length and structure - Every distractor based on a real misconception or common error — no throwaway options - Apply+ questions must use a novel scenario, not just restate the textbook - Distribute correct answer positions roughly evenly across A-D **STUDENT VERSION** Quiz title, name/date line, instructions, numbered questions with A-D choices. No Bloom's labels visible. Format: {{{format: print-ready, Google Forms, LMS import}}} **TEACHER ANSWER KEY** Per question: - Correct answer + Bloom's level - Why it's correct (1-2 sentences) - Per distractor: "A student who picks [X] likely thinks..." (the diagnostic value) **After the quiz:** - Bloom's distribution confirmation - Answer position spread (A/B/C/D count) - Top 2-3 re-teach priorities if students cluster on specific distractors Rules: Only test what's been taught. Distractors diagnose, never trick. Higher-order means a cognitive move beyond recall, not harder words. No grammatical clues or cultural bias in scenarios.