Complete AI Training
Sign inGet my AI kit

Your job's AI kit

Get your AI kit

Tell us who you are and what you do. We show you your kit right away and email you the link: skills, prompts, AI agents, MCP servers and courses for your job.

500+ jobs ready, and we make a kit for any other job. No payment needed to look.

Share

AI agent for economists

Peer Review Replication Agent

Reproduce a paper's main results or explain exactly why they differ.

AnalysisAutomationFinanceScience and research
See how it worksGet access
Peer Review Replication Agent: what goes in, what the agent does and what you get

What it does

Replicating a paper means rebuilding the data, rerunning the model and checking that the coefficients match, which can take weeks. This agent reads the paper and its replication files and lists the data, sample rules, variables and model specification. It rebuilds the dataset, runs the model and compares each coefficient and standard error with the published table. When something differs, it investigates in order: sample size, variable definitions, data versions and software options, rerunning after each change. It stops when it matches or all causes are explored, and writes a replication note listing what matched. The economist approves the conclusions before anything is shared.

How it works

Follow the arrows from top to bottom. The orange dashed arrow is the loop: when a check fails, the agent goes back and tries again.

Start and resultWhat it doesA check on its own workWaits for your OKGoes back and retries
Yes, continueYes, continueApprovedNoNo 1 STARTS WHEN Paper assigned for replication 2 USES A TOOL Read the paper and list data, sample rules and themodel 3 USES A TOOL Rebuild the dataset from the files 4 CHECKS THE RESULT Does the sample size match the paper? If not: review sample rules and missing data handling,then rebuild. Back to step 3. 5 USES A TOOL Run the model 6 DOES Compare coefficients and standard errors with thepaper 7 CHECKS THE RESULT Do all main results match within rounding? If not: test data versions, variable definitions andsoftware options one by one. Back to step 3. 8 DOES Record each difference and its cause 9 DOES Write the replication note 10 YOU APPROVE Economist approves the conclusions 11 RESULT Replication note with tables
Read the steps as a list
  1. Paper assigned for replication
  2. Read the paper and list data, sample rules and the model
  3. Rebuild the dataset from the files
  4. Does the sample size match the paper?If not: review sample rules and missing data handling, then rebuild. Back to step 3.
  5. Run the model
  6. Compare coefficients and standard errors with the paper
  7. Do all main results match within rounding?If not: test data versions, variable definitions and software options one by one. Back to step 3.
  8. Record each difference and its cause
  9. Write the replication note
  10. Economist approves the conclusionsThe agent waits here for your OK.
  11. Replication note with tables

How it decides

It treats a result as replicated when coefficients agree within rounding, and otherwise tests listed causes in order of likelihood.

  • Count a coefficient as matching within 1 percent
  • Test sample rules before variable definitions
  • Try the stated software version before others
  • Report unexplained differences as unresolved, not as errors

Make it yours

Every agent is a starting point. You choose these settings for your own situation.

  • Matching tolerance (default 1 percent)
  • Software and version
  • Tables to replicate
  • Note format

What keeps you in control

It always asks you first

  • Economist approves the conclusions
  • Economist approves contact with the paper's authors

Hard limits

  • Never accuses authors of error
  • States every difference with evidence

It stops when

  • Done: results match or all causes are explored and documented
  • Stop: data files are not available

Set it up

We guide you through the set-up, step by step

Members get the full set-up guide for this agent. No technical skills needed: you copy, paste and upload.

10 minto set it up in your AI
5 AIsChatGPT, Claude, Copilot, Gemini, Grok
  • One set of instructions to paste into your AI, with the clicks for ChatGPT, Claude, Microsoft 365 Copilot, Gemini and Grok
  • The agent then walks you through connecting your own data, one source at a time
  • A downloadable copy with the flow chart, the rules and the full guide
Get access to this agent

An example run

What happensThe paper reported 2,412 observations and a coefficient of 0.31 on trade exposure. The agent first got 2,466 observations. The sample check failed, and the paper excluded firms with missing sales, which it applied. The sample then matched at 2,412 and the coefficient was 0.31, but one standard error was 0.09 instead of 0.07, traced to clustering. The economist approved the note.

More agents for economists

Budgets and money · Decisions

Budget Scenario Optimization Agent

A small set of balanced, rule-compliant budget scenarios ranked by how well they meet priorities

10 steps2 checks3 jobs
Forecasting · Government

Economic Forecast Refresh Agent

A refreshed forecast with explained changes and a short note ready for decision makers

10 steps2 checks6 jobs
Reporting · Creating

Regional Economic Brief Builder Agent

Produce a consistent, checked regional brief in hours instead of days.

11 steps2 checks2 jobs
Quality control · Analysis

Survey Weighting and Sample Check Agent

Produce weighted survey estimates that are stable and well documented.

10 steps2 checks5 jobs