domains / families / research / jqte-sensitivity

jqte-sensitivity

separate

Use to design and report robustness for a 《数量经济技术经济研究》 (JQTE) manuscript where the contribution is measurement, a model, or a forecast — focusing on sensitivity to method choice, parameter / specification, data caliber, and sample window, rather than the placebo / parallel-trends battery of a causal paper. Use after the main estimates exist.

When to use

• Sensitivity to method choice
• Sensitivity to parameter or specification
• Sensitivity to data caliber
• Sensitivity to sample window
• Robustness for measurement contributions
• Robustness for model contributions
• Robustness for forecast contributions

When not to use

• Placebo tests
• Parallel trends tests
• Randomization tests
• Causal paper robustness batteries

Skill page

Source