M5: Decision Validation Method - Selection Logic

Overview

Validate decisions via follow-up metrics (fit, regret, and consistency) and iterate.

Why validation matters (zh-aligned)

Without validation, methods cannot improve (A3) and claims cannot be falsified. Validation turns theory into measurable practice.[^1]

Validation protocol (template)

1. Pre-register needs and criteria (M1/M2 artifacts).
2. Define a time window (7 days / 30 days).
3. Measure:
   - regret score,
   - need-consistency,
   - fit score stability,
   - switching/return behavior (when applicable).
4. Write a short post-mortem: what signals misled you?
5. Update your heuristics or weights for next time.

References

1. Popper, K. R. (1959). The Logic of Scientific Discovery. Routledge. (Original work published 1935)[source]
2. Kahneman, D. (2011). Thinking, Fast and Slow. Farrar, Straus and Giroux.[source]

Further Reading

• A3: Improvability
• T3: Consistency improvement
• Theory validation process