Appendices

This section collects supplementary material, extended analysis, and design reflections referenced throughout the main text. While not required for core understanding, these appendices offer deeper insight into decisions made during the development of FallacyTag.

• Appendix A is the transcript of GPT-4’s logical check of Veritassium’s “Why Democracy Is Mathematically Impossible” video.
• Appendix B is the transcript of GPT-4’s logical check of a few posts on Reddit CMV on Local vs National Politics.
• Appendix C discusses the challenges and limitations of using user feedback to improve the system.
• Appendix D presents a structured view of the FallacyTag architecture as a product line architecture (PLA)
• Appendix E explains why the seven scorecard dimensions are scored independently and weighted equally. It discusses the risks of double-counting and meta-debates, acknowledges natural correlations, and outlines when more sophisticated weighting might be justified in future deployments.

---

Table of contents

• Appendix A - Logic Check Experiment: Why Democracy Is Mathematically Impossible by Veritasium
• Appendix B - Logic Check Experiment: Reddit CMV on Local vs National Politics
• Appendix C: On Learning from User Feedback
• Appendix D: FallacyTag Variability Model
• Appendix E - Independence, Correlation, and Weighting