Skip to content

README.md

Latest commit

 

History

History
34 lines (28 loc) · 3.4 KB

README.md

File metadata and controls

34 lines (28 loc) · 3.4 KB
Raw

AEJ: Applied — Resources

Capability layer for the American Economic Journal: Applied Economics skill pack. The runnable code is vendored for self-containment; the cross-cutting method guidance lives once in the shared empirical-methods hub and is linked below.

Contents

Resource What it gives an agent
code/ Reproducible Stata + Python causal-inference skeleton (clean → descriptive → DiD/IV/RDD/DML → mechanism → robustness → tables). Vendored from shared-resources/empirical-methods/code (Stata 18 MP, 2026-06); copy-and-adapt as the backbone of your openICPSR replication package — change no command blindly.
worked-examples/01-introduction.md See a before→after rewrite of a paper introduction in AEJ: Applied house style (question → credible identification → headline causal estimate with SE → mechanism → policy/economic lesson). Illustrative fictional applied-micro paper.
exemplars/library.md Benchmark against real, web-verified AEJ: Applied papers (confirmed 10.1257/app. DOI stem) organized by method × topic. Design positioning only — no fabricated numbers; includes a sibling-confusion guard vs. AER / AEJ: Economic Policy / field journals.
reviewer-objection-checklist The objections referees actually raise, by identification strategy (DiD / IV / RDD / DML / matching / mechanism), each with its preemption. Stress-test the design before drafting.
reporting-standards Modern inference + reporting table stakes: SE clustering, weak-IV diagnostics, multiple-testing, DiD/RDD reporting, reproducibility.
execution-with-mcp Guidance → a fitted, audited result. Maps each design family / reviewer objection to the concrete StatsPAI / Stata MCP tools in this environment, so this pack's identification / analysis skills run the modern DiD / weak-IV-robust CI / multiple-testing correction and report the number.
official-source-map.md Venue-specific facts (scope, editors, fee, single-blind review, data policy, house citation style) with sourcing discipline.
external_tools.md External tools / packages relevant to this venue.

How to use

  1. Before drafting tables — run the design against the reviewer-objection checklist; every objection you cannot answer in one paragraph + one exhibit is a hole (see ../skills/aeja-referee-strategy/SKILL.md).
  2. When building results — adapt code/ to your data; keep the modern-estimator chain (heterogeneity-robust DiD, weak-IV-robust IV, bias-corrected RDD), and grow it into the openICPSR package the AEA Data Editor will check (../skills/aeja-replication-package/SKILL.md).
  3. Before submission — walk the reporting-standards inference audit and this pack's official-source-map.md for venue rules (../skills/aeja-submission/SKILL.md).

Method guidance here is venue-neutral; always defer to this pack's skills and official-source-map.md for what this journal specifically requires.