AgentSkillsCN

analyzing-survey-results

解读调研数据，提炼可落地的洞察与建议。适用于处理问卷回复、撰写调研报告，或将量化分析结果转化为切实可行的行动方案时使用。

SKILL.md
--- frontmatter
name: analyzing-survey-results
description: Interpret survey data and extract actionable insights. Use when processing survey responses, writing survey reports, or translating quantitative findings into recommendations.

Analyzing Survey Results

Quick start

Collect or infer:

  • •Survey data (responses, sample size, collection method)
  • •Survey design (question types, scales, skip logic)
  • •Business context (what decisions the survey should inform)
  • •Audience for findings (executives, product, research)

Then produce output using TEMPLATES.md. Validate with RUBRIC.md.

Workflow

  1. •Assess data quality (response rate, completion rate, sample representativeness)
  2. •Analyze closed-ended questions (frequencies, means, distributions)
  3. •Identify significant patterns and segments
  4. •Analyze open-ended responses (theme coding)
  5. •Synthesize findings with data quality caveats
  6. •Write recommendations tied to specific data points
  7. •Run the rubric check. Revise until it passes.

Degrees of freedom

  • •Low: Statistical accuracy, data quality reporting
  • •Medium: Emphasis and prioritization of findings
  • •High: Narrative framing, recommendation specificity

State awareness

  • •If sample size is small (<100): report directional findings, avoid percentages implying precision
  • •If response rate is low (<20%): flag non-response bias risk
  • •If scale data: report means with distribution shape
  • •If NPS/CSAT: include benchmarks for context
  • •If segments differ significantly: lead with segment analysis

Failure modes to avoid

  • •Presenting percentages without sample sizes
  • •Over-interpreting small differences as meaningful
  • •Ignoring non-response and selection bias
  • •Cherry-picking data that supports a narrative
  • •Presenting findings without actionable recommendations
  • •Missing key segment differences

References