360 Secure SkillHub
360 安全技能中心

Quiz

该 Skill 提供测验设计指南,涵盖知识测评、性格测验、评估、获客与问答等类型;包含情境加载、通用规则、题型参考、出题检查清单、评分模式、互动增强与常见风险,并可按需加载类型、出题与实现参考文件。

ivangdavila
ivangdavila
热度 89.05
最近更新 2026-05-11 16:28:08
分类:知识依赖:纯提示词应用场景:测验 / 教学辅导 / 文案创作

安装技能

复制下面的话术发给你的 Agent,它会通过 secure-skills 检索并安装该技能。

执行命令安装此 skill: npx secure-skills-cli add quiz
Metadata
name
Quiz
description
Design engaging quizzes with effective questions, scoring logic, and results that drive learning or conversions.

Situation Detection

Context Load
Knowledge assessment, exams, certifications types.md → Knowledge section
Personality quizzes, "Which X are you?" types.md → Personality section
Lead generation, marketing quizzes types.md → Lead-gen section
Writing effective questions questions.md
Building quiz UI/UX, gamification implementation.md

Universal Rules

One concept per question. Double-barreled questions confuse and measure nothing. "Do you like pizza and exercise?" → Bad.

Wrong answers must be plausible. If correct answer is obvious by elimination, you're testing pattern recognition, not knowledge.

Results must feel personal. Generic outcomes kill engagement. "You got 7/10" loses to "You're an 80s Movie Expert — you caught references most people miss."

Progress visibility motivates. Show question count, progress bar, time remaining. Uncertainty creates anxiety and abandonment.


Quiz Types Quick Reference

Type Goal Typical Length Results
Knowledge Assess learning 10-20 questions Score + feedback per answer
Personality Engagement, sharing 5-12 questions Personality type/category
Assessment Diagnose level/fit 10-30 questions Detailed report
Lead-gen Capture email 5-8 questions Results gated behind email
Trivia Entertainment Any Leaderboard, social share

Question Design Checklist

  • Clear, unambiguous wording
  • One correct answer (or explicit multi-select instruction)
  • Distractors are plausible, not obviously wrong
  • No "all of the above" or "none of the above" (lazy design)
  • Avoid negatives ("Which is NOT...")
  • Test the concept, not reading comprehension

Scoring Patterns

Simple percentage: Correct/total × 100. Best for knowledge tests.

Weighted scoring: Some questions worth more. Good for prioritized competencies.

Branching outcomes: Answer combinations map to results. Used in personality quizzes.

Diagnostic rubric: Score across multiple dimensions. Best for assessments and skill evaluations.


Engagement Boosters

  • Immediate feedback after each answer (right/wrong + explanation)
  • Visual progress indicator
  • Streak rewards ("3 in a row!")
  • Time pressure (optional, increases excitement but also anxiety)
  • Social sharing of results
  • Leaderboards for competitive contexts

Red Flags

  • All correct answers in position B/C → Detectable pattern
  • Questions testing obscure trivia vs actual learning objectives
  • Results that don't connect to actions ("Now what?")
  • Too long with no progress indication → Abandonment
  • Mobile-unfriendly UI (tiny buttons, horizontal scroll)

When to Load More

Situation Reference
Designing for specific quiz type types.md
Writing and reviewing questions questions.md
Building quiz flow, UI, tools implementation.md