domains / families / research / restud-robustness

restud-robustness

separate

Use when the main results of a The Review of Economic Studies (REStud) manuscript exist but referee-anticipating checks — robustness, heterogeneity, mechanism, placebo, alternative specifications — are missing or fragile. Hardens the result against demanding referees; does not redesign identification.

When to use

• Specification robustness
• Placebo / falsification tests
• Inference robustness
• Mechanism evidence
• Heterogeneity analysis
• Selection / attrition / measurement checks
• Coefficient-stability tables
• Stress-testing fragile assumptions

When not to use

• Redesigning identification
• Robustness theater
• Data-mining subgroups

Skill page

Source