agentsclimarketplace

02 paper strategy

Skill lianxhcn/PXa2026a/skills/core/02-paper-strategy

Lecture

Install
npx -y skills add lianxhcn/PXa2026a --skill 02-paper-strategy

Assembled from the repository path, not quoted from the project. Check it against their README if it does not work.

3 things to look at

  • 20 days oldThe repository was created 20 days ago. New is not bad, but a brand new repository carrying a familiar-sounding name is the shape a typosquat arrives in, and there has been no time for anyone else to find a problem with it.
  • no licenseNo license file was found in the repository. Code published without one is not open source by default, so using it at work is a question for whoever answers licensing questions where you are.
  • 3 stars3 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.

What its author says it does

Copied from the file, not written here

Use this skill to break down an empirical paper's identification strategy — its research question, setting, data, econometric method, and identification/endogeneity handling (a Problem-Setting-Data-Method-Identification decomposition). Explains what each method is for and why the author uses it, and maps key figures/tables to methods and Stata/R commands. Stays at the concept level (no formula derivations); never fabricates.

SKILL.md

3.0 KB, as published. Nobody here has run it

paper-strategy(论文实证策略拆解)

实证方法与识别角度拆解一篇论文:它用了哪些方法、凭什么可信、怎么处理内生性。按 P-S-D-M-I 五层梳理——问题(Problem)、情境(Setting)、数据(Data)、方法(Method)、识别(Identification)——并给出图表 → 方法 → 命令对照。对应第 1 讲配套阅读笔记的「实证方法与识别」,也用于第 3 讲。

何时用

  • 想弄清一篇论文用了什么计量方法、识别策略是什么、怎么应对内生性;
  • 想判断一篇论文的方法能不能迁移到自己的研究;
  • 想顺着关键图表,快速建立「图表对应什么方法、方法对应什么命令」的对照。

使用方式(提示词)

把论文 PDF 连同下面的要求发给 agent:

只从「实证方法与识别」这一个角度帮我写一份中文笔记,面向计量基础较弱的读者,包含:

  • 核心方法:主要用了哪些计量方法?每种一两句话说清它是干什么用的,不要写公式推导
  • 识别策略:凭什么相信估计的是因果效应?处理组、对照组、政策时点分别怎么来的?
  • 内生性与稳健性:作者担心哪些干扰(趋势不平行、选择偏差、溢出污染等),又怎么缓解?
  • 图表与方法与命令对照:挑几张关键图表,说明它用的核心方法与对应的 Stata/R 命令。

只依据论文本身,不编造;名词保持克制;每条结论标注对应的图或表编号。

输出约定

  • 五层结构(P-S-D-M-I)+ 一张「图表 / 方法 / 命令」对照表;
  • 守初级班难度红线:只讲方法「是什么、为什么用」,不做公式推导;超纲名词点到为止;
  • 标注图表编号,便于回原文核对。

最小使用示例

用 paper-strategy 拆一下这篇论文的识别策略,重点告诉我它怎么处理内生性,以及主结果那张图用的什么方法、什么命令。

人工检查清单

  • AI 辅助、不替代判断:方法能否迁移、识别是否站得住,由研究者自己把关。
  • 图表编号、命令名回原文与复现代码核对。
  • 超出你当前掌握的方法(如双重稳健、三重差分),先记下来即可,后续讲次与高级班再系统学。

来源与许可

  • 原创:本课程为「连享会 2026 暑期班·初级班」编写,Agent 无关的开放 Skill,MIT(随仓库 LICENSE-CODE)。
  • 最后验证日期:<!-- TODO: T7 彩排双端实测后填写 -->

Keep looking

Skills are one crate of 328,083. Ordering is by how many stacks a row turns up in, so the top of any crate is what has actually been picked rather than what has the most stars.