科研海报图表与表达审查
审查科研海报的图表编码、尺度与结论措辞,输出误读风险清单及证据表达改进项。
完整提示词
Main objective: > Perform a rigorous, transparent and reproducible workflow for Research Poster Design, with a clear audit trail for all methodological decisions. ## 1. CONTEXT Define question, unit of analysis, data/material, source/provenance, time range, software/tools, transformations, decision stakes and intended claim. ## 2. DOMAIN STANDARD Scientific communication must preserve scale, denominator, uncertainty and distinction between result and interpretation. Visual design should improve comprehension without altering the evidentiary meaning. ## 3. SPECIALIZED FOCUS For **Research Poster Design**: - define the exact artifact, method or claim being audited - preserve raw evidence and provenance - distinguish planned from post hoc choices - make transformations and exclusions reproducible - test plausible alternatives - state unresolved uncertainty ## 4. INTEGRITY CHECKS - axes and scales are defensible - denominators are explicit - uncertainty is visible where material - color/encoding is accessible - figure/table can stand alone with caption - results are separated from discussion - abstract matches full results - public summary does not upgrade certainty ## 5. AUDIT CARD ```text Object: Source/provenance: Method: Transformation: Assumption: Evidence: Quality issue: Alternative: Reproducibility status: Decision impact: ``` ## 6. REQUIRED MATRICES ### Figure/Table Integrity Matrix | Element | Data represented | Encoding | Scale | Uncertainty | Misread risk | Fix | |---|---|---|---|---|---|---| ### Claim Communication Matrix | Claim | Evidence | Audience wording | Uncertainty | Forbidden overstatement | |---|---|---|---|---| ## 7. ANTI-FAILURE RULES Do not hide exclusions, overwrite raw data, infer saturation from convenience, confuse coding frequency with importance, truncate axes deceptively, suppress uncertainty, or use polished visualization to imply stronger evidence than exists. ## 8. REQUIRED OUTPUT 1. Scope and provenance. 2. Method and assumptions. 3. Quality/integrity findings. 4. Required matrices. 5. Sensitivity / alternative choices. 6. Reproducibility package requirements. 7. Inference/communication limits. 8. Remediation priorities. End with **Scientific Integrity Check** confirming traceability from raw evidence through transformation to final claim. This supports scientific analysis and does not replace domain-specific ethical or institutional procedures.
分类与标签
设计创意、科研海报、图表审查、科学可视化、证据表达