Audit impact evaluation reports for causal inference validity.
npx skills add https://github.com/amnotyoung/dev-eval-agents --skill impact-review
Please help me install this Agent Skill. Skill: impact-review Source: https://github.com/amnotyoung/dev-eval-agents/tree/main/skills/impact-review Command: npx skills add https://github.com/amnotyoung/dev-eval-agents --skill impact-review
This Skill addresses the lack of standardized methodological rigor in impact evaluation reports by providing a structured audit framework based on causal inference principles.
Use the impact-review skill to audit the methodology of the impact evaluation report located in the current project directory.
High-intent search queries and answers about installing and using this skill.
Auditing an impact evaluation report involves applying a structured 5-axis, 10-question framework to evaluate causal identification strategies like PSM, DiD, and RCT, determining if the methodology is suitable, conditionally acceptable, or unsuitable.
A counterfactual design in causal inference establishes what would have occurred without an intervention, serving as the core mechanism for verifying statistical robustness and project accountability in ODA impact evaluations.
Yes, you can use this methodological audit for ODA project assessments because it strictly adheres to KOICA Evaluation Guidelines and specific causal identification criteria to ensure evidence-based project accountability.
The 5-axis assessment framework evaluates statistical robustness by applying 10 targeted questions to an impact evaluation report, verifying the validity of the counterfactual design and causal identification strategies.
You need to verify causal identification strategies in an impact evaluation when reviewing a development assistance project report to ensure methodological validity and a sound statistical approach before final human approval.
A limitation of using causal inference frameworks for project accountability is that the audit determines methodological suitability but still requires final human approval to fully validate evidence-based project accountability.
amnotyoungEvaluate ODA project evaluation reports against KOICA Evaluation Guidelines v2.
zhnnky329Run robustness, sensitivity, error, and baseline checks on modeling outputs.
nahisahoEstimate causal effects from observational data using PSM, IPW, DID, and RDD.
Likas07Synthesize user feedback from multiple channels into actionable product insights.
