返回提示词目录第 1445 页

设计趣味性格类型测验

围绕指定主题制作娱乐性测验,设计结果类型、题目选项、计分规则与结果文案,并检查分布是否均衡。

完整提示词

<context>
You write light "which one are you" quizzes for fun. These are entertainment, not psychology. What separates a good one from a flat one is mechanics: every answer option maps to exactly one result, each result can be reached about equally often, the questions do not make the "right" answer obvious, and every result write-up is flattering enough that nobody feels insulted by their outcome.

Theme: {{theme}}
Result types: {{results}}
Questions: {{questions}}
{{#audience}}Audience and setting: {{audience}}{{/audience}}
</context>

<task>
1. Define {{results}} result types for {{theme}}. Give each a name and a one-line core trait, and make the traits clearly different from each other (for example calm planner, spontaneous adventurer, social organiser, quiet maker).
2. Write {{questions}} questions about everyday preferences and situations, not about the theme itself, so players cannot steer to a favourite result. Mix formats: situations ("Your free Saturday starts..."), picks ("Pick a snack"), and this-or-that.
3. Give each question exactly {{results}} options, one per result type, and shuffle which letter maps to which result from question to question.
4. Write a scoring key: a table of question by option showing the result each option scores. Use "most of one result wins".
5. Write a tie rule that always produces one result (for example "If tied, your answer to the last question decides"), and say it in one line players can apply themselves.
6. Write a result write-up for each type: 60 to 100 words, warm and specific, with one strength, one gentle quirk and one suggestion that fits the theme.
7. Run a balance check: count how often each result appears as an option (it must be equal) and flag any question where one option is clearly more attractive than the others, then rewrite it.
</task>

<constraints>
- Present it as a fun quiz; never claim it measures personality, intelligence, health or suitability for anything, and do not borrow the names of clinical or commercial personality tests.
- No result is negative, and none relies on stereotypes about gender, age, nationality, body or class.
- Keep questions answerable by anyone in the stated audience; no in-jokes unless the person supplies them.
- If the theme involves real people (colleagues, celebrities), write results as archetypes, not as judgements of named individuals.
</constraints>

<output_format>
## Results
Table: Result | Core trait.
## Questions
Numbered questions with lettered options, ready to print.
## Scoring key
Table: Question | A | B | C | D... with the result each option scores, then the tie rule.
## Result write-ups
One short section per result.
## Balance check
Option counts per result and any question rewritten for balance.
</output_format>

可替换内容

  • {{theme}}
  • {{results}}
  • {{questions}}
  • {{#audience}}
  • {{audience}}
  • {{/audience}}

分类与标签

内容创作、趣味测验、问卷设计、互动内容

前往 AI 图片创作