Menu
执行双重差分法(DID)因果推断分析，结合计量经济学理论、统计检验方法、政策评估实践和数据科学技术。包含实验设计、模型设定、因果估计、政策解释四个阶段。
The review-first prompt is selected by default. You can switch to a direct command or download a local copy.
Read SKILL.md and any companion files shown by SkillsMP before deciding whether to install.
Based on SOC occupation classification
Install with Codex or Claude Copy this prompt, paste it into Codex, Claude, or another assistant, and let it review the skill page and install it for you.
A direct command skips the review prompt. Inspect the source before running it.
npx skills add https://github.com/ptreezh/sscisubagent-skills --skill did-analysisThe command stays on one line. Scroll horizontally to inspect it before copying.
Prefer a local copy? Download the files currently available to SkillsMP.
Showing SKILL.md
| name | did-analysis |
| description | 执行双重差分法(DID)因果推断分析，结合计量经济学理论、统计检验方法、政策评估实践和数据科学技术。包含实验设计、模型设定、因果估计、政策解释四个阶段。 |
| license | MIT |
| compatibility | Python 3.8+, pandas, numpy, scipy, statsmodels, matplotlib, seaborn |
| metadata | {"domain":"causal-inference","methodology":"difference-in-differences","complexity":"advanced","version":"2.0.0","integration_type":"theoretical_quantitative","author":"socienceAI.com","website":"http://agentpsy.com"} |
| allowed-tools | python bash read_file write_file task |
prompts/experimental-design.mdprompts/model-specification.mdprompts/causal-interpretation.mdprompts/policy-recommendation.mdscripts/did_estimator.pyscripts/parallel_trend.pyscripts/robustness_test.pyscripts/visualization.pyscripts/integrated_did.py# DID集成分析器的核心决策逻辑 - 五阶段研究流程
class IntegratedDIDAnalyzer:
def execute_analysis(self, data, policy_context):
# 阶段1: AI进行实验设计 (理论指导)
experimental_design = self.execute_experimental_design(policy_context)
# 阶段2: AI进行模型设定 (理论+方法)
model_specification = self.execute_model_specification(data, experimental_design)
# 阶段3: 脚本执行计量估计 (精确计算)
estimation_results = self.execute_did_estimation(data, model_specification)
# 阶段4: AI进行因果解释 (理论阐释)
causal_interpretation = self.execute_causal_interpretation(estimation_results, experimental_design)
# 阶段5: AI制定政策建议 (实践应用)
policy_recommendations = self.execute_policy_recommendation(causal_interpretation, policy_context)
return self.integrate_results(experimental_design, estimation_results, causal_interpretation, policy_recommendations)
experimental-design.mdmodel-specification.mdcausal-interpretation.mdpolicy-recommendation.mdAI任务: 加载experimental-design.md，进行深度政策机制分析
输出: 实验设计方案、组别选择、时间窗口设定
计量支持: 数据特征分析脚本提供数据描述
AI任务: 加载model-specification.md，设计DID模型框架
输出: 模型设定方案、内生性处理策略、异质性分析设计
计量支持: 计量理论基础和模型选择指导
AI任务: 决策调用哪些计量脚本 脚本执行:
did_estimator.py: 执行DID估计parallel_trend.py: 平行趋势检验robustness_test.py: 稳健性检验
输出: 精确的估计结果、统计检验、质量指标AI任务: 加载causal-interpretation.md，进行因果机制阐释
输出: 因果效应解释、理论意义、统计可靠性评估
计量支持: 估计结果提供解释基础
AI任务: 加载policy-recommendation.md，制定政策建议
输出: 政策建议、可行性分析、外推有效性评估
计量支持: 因果效应大小和政策含义
每个提示词文件都采用渐进式结构：
理论质量: 计量模型的理论一致性和识别策略有效性
统计质量: 平行趋势假设、统计显著性、估计精度
政策质量: 政策现实相关性、建议可行性、外推有效性 if not check_parallel_trend_assumption(did_results): quality_issues.append("平行趋势假设不满足")
if not assess_endogeneity_concerns(did_results): quality_issues.append("存在潜在内生性问题")
if not validate_robustness_results(did_results): quality_issues.append("稳健性检验未通过")
return quality_issues
## 开始使用
### 快速启动
```bash
# 进入技能目录
cd did-analysis/
# 运行集成分析
python scripts/integrated_did.py
prompts/experimental-design.md: 实验设计专家提示prompts/model-specification.md: 模型设定专家提示prompts/causal-interpretation.md: 因果解释专家提示prompts/policy-recommendation.md: 政策建议专家提示scripts/did_estimator.py: DID估计核心算法scripts/parallel_trend.py: 平行趋势检验scripts/robustness_test.py: 稳健性检验scripts/visualization.py: 数据可视化scripts/integrated_did.py: 集成分析引擎references/METHODOLOGY.md: DID方法论详解references/BEST_PRACTICES.md: 最佳实践指南references/COMMON_PITFALLS.md: 常见陷阱与避免assets/templates/: 数据模板和报告模板注意: AI负责理论思考和因果解释，脚本负责精确计量估计，通过智能决策引擎实现因果推断的科学性和政策相关性的统一。