agentsclimarketplace

Jebo replication package

Skill brycewang-stanford/Awesome-Journal-Skills/Journal-of-Economic-Behavior-and-Organization-Skills/skills/jebo-replication-package

Use when assembling the data, code, and experimental-materials deposit for a Journal of Economic Behavior & Organization (JEBO) manuscript — z-Tree/oTree code, instructions, raw and analysis data, and an Elsevier Research Data / Mendeley Data statement. Builds a reproducible package; it does not run the analysis.From its SKILL.md

Install
npx -y skills add brycewang-stanford/Awesome-Journal-Skills --skill jebo-replication-package

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

SKILL.md

4.9 KB, 999 tokens by cl100k_base, as published. Nobody here has run it

Replication & Materials Package (jebo-replication-package)

When to trigger

  • The analysis is settled and you must assemble what others need to reproduce it
  • An experiment is done but the instructions and z-Tree/oTree code are not yet packaged
  • A referee or editor asks where the data and code are
  • You are writing the Elsevier "data availability" / Research Data statement
  • An agent-based model's runnable code and seeds are not yet organized for sharing

JEBO's reproducibility posture

JEBO follows the Elsevier Research Data policy: data and code sharing is strongly encouraged (via Mendeley Data or a linked repository) but not currently a hard pre-acceptance gate the way the AEA Data Editor or the Econometric Society's pre-acceptance check are (检索于 2026-06;以官网为准). In practice, for an experimental journal, referees increasingly expect the full materials, and a credible deposit strengthens acceptance. Treat the package as part of the contribution, not an afterthought.

The distinctive JEBO requirement is experimental reproducibility: a reader must be able to re-run your experiment, not just your regressions. That means the instructions, the screen-by-screen protocol, and the experiment software are first-class deposit items.

What the package must contain

ItemDetail JEBO referees look for
Experiment instructionsfull participant-facing text (translated to English if run in another language), including comprehension questions
Experiment softwarez-Tree .ztt / oTree app code, or the platform code, runnable with setup notes
Screenshots / protocolsession protocol, screen flow, randomization procedure, payment scheme
Raw datasubject-level raw output (anonymized), with a codebook
Analysis codescripts that go from raw data to every table/figure, with a master run file
Software environmentversions (Stata/R/Python, z-Tree/oTree), packages, seeds for any simulation
READMEone-command (or clearly-ordered) reproduction path; maps each script to each exhibit
Pre-registration linkregistry URL + a note on deviations, if pre-registered
Ethics / consentIRB/ethics approval reference; consent procedure

For agent-based / simulation papers, the deposit is the model code, parameter files, seeds, and the sweep scripts that regenerate the reported figures.

Privacy and anonymization

Strip direct identifiers from raw subject data; for field data, follow consent and data-protection constraints and document any access restrictions. State clearly if data cannot be shared and why (Elsevier allows a justified restricted-access statement).

The data-availability statement

Write the Elsevier "Data availability" statement to match reality: a Mendeley Data DOI / repository link, or a justified statement of restriction. Vague statements ("data available on request") are weak and increasingly disfavored; deposit where possible.

Checklist

  • Experiment instructions (English) + comprehension questions included
  • z-Tree/oTree or platform code included and runnable, with setup notes
  • Session protocol, screen flow, randomization, and payment scheme documented
  • Raw subject-level data (anonymized) + codebook
  • Analysis scripts reproduce every table/figure from raw data via a master file
  • Software versions, packages, and seeds recorded
  • README maps each script to each exhibit; reproduction path is explicit
  • Pre-registration link + deviations noted (if applicable); IRB/consent referenced
  • Data-availability statement points to a real deposit (Mendeley Data) or a justified restriction

Anti-patterns

  • Sharing analysis code but not the experiment software/instructions (cannot re-run the experiment)
  • "Data available on request" with no deposit
  • Raw data with direct identifiers left in
  • A README that lists files but never gives the reproduction order
  • Simulation results with no seeds or sweep scripts
  • Claiming pre-registration without linking the registry or disclosing deviations

Output format

【Deposit location】Mendeley Data DOI / repository link / justified restriction
【Experiment materials】instructions + z-Tree/oTree code + protocol [Y/N]
【Data】raw (anonymized) + codebook [Y/N]
【Code】master run file → every exhibit; versions + seeds recorded [Y/N]
【README】explicit reproduction path mapping script→exhibit [Y/N]
【Pre-reg + ethics】registry link + deviations; IRB/consent [Y/N / n.a.]
【Next step】jebo-referee-strategy

What ships with it

Read from the repository

Just SKILL.md. No reference files, no scripts.

Keep looking

Skills are one crate of 325,949. 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.