Selected decisions that shape the method.
This register explains the main choices behind the method.
Available for research; independent replication pending. ArtiProof is designed to support evidence-based assessment of organisational knowledge. It is not yet broadly independently validated. Reproducibility between independent assessors and proportionality in practice remain open questions; the first independent replication remains pending.
DR-VAL-CAL-01 — Calibration before complexity
When assessor disagreement occurs, improve reproducibility first through calibrated examples, clearer evidence guidance, sharper scope and transparent reasoning before adding normative criteria, scores or assurance properties.
The Assessor Calibration Library is methodological support rather than part of the normative core.
Related design consequences
- no aggregate maturity score is added to hide disagreement;
- the current five properties remain separately visible;
- material disagreement should be recorded before consensus;
- experiments must be capable of showing that the existing architecture is insufficient.
Boundaries of public claims
Accepted principle
Public material must distinguish properties the method is designed to support from properties demonstrated by empirical evidence. Reproducibility between independent assessors and proportionality of additional effort remain validation claims. Publication as research-ready does not establish either result.