-
Notifications
You must be signed in to change notification settings - Fork 0
Open
Description
This checklist is designed for a generic LLM acting as a senior reviewer of tables and quantitative results in a scientific paper. Each item can be treated as a step in an automated review workflow.
1️⃣ Information Presence Check (Structure)
- Are all relevant metrics reported for the study type?
- Are descriptive statistics included (e.g., mean, SD, CI)?
- Are sample sizes / n per group clearly reported?
- Are any essential metrics missing that a Q1 journal would expect?
- Are units, definitions, and thresholds clearly specified?
2️⃣ Value Plausibility Check
- Are the reported values internally consistent across metrics?
- Are the effect sizes or differences plausible given the underlying study design?
- Are the uncertainty estimates (SD, CI, p-values) reasonable?
- Are there any signs of overfitting, deterministic behavior, or simulation artifacts?
- Are additional metrics or alternative statistics needed to improve robustness?
3️⃣ Overall Acceptance Assessment
- Based on the table(s), assign a preliminary acceptance likelihood:
- ⬜ Acceptable
- ⬜ Borderline / minor revision needed
- ⬜ Likely to be rejected
- Document the main reasons for this assessment.
4️⃣ Improvement Recommendations (Step-by-Step)
- Essential fixes: critical issues that must be addressed for acceptance
- Optional enhancements: suggested improvements to strengthen rigor and clarity
- Recommendations should include:
- improving statistical robustness
- clarifying metric definitions
- increasing transparency and interpretability
- aligning analysis with Q1 journal standards
5️⃣ Review Style
- Maintain a critical but constructive tone
- Be explicit, technical, and precise
- Avoid vague or generic statements
- Assume the audience is experienced researchers
Metadata
Metadata
Assignees
Labels
No labels