EM-PEO-06 · Subject model · W3
Performance, Evidence and Calibration
The review cycle, criteria, observations, feedback, achievements, disputes and calibration decisions. Assessments must not become permanent properties of a person.
Partially published; research continues
Claude: complete-browser-memo-and-semantic-text-audit; Grok: complete-browser-research-memo.
Published results
Research note, in Russian: Full contour unfinished: disposition shared scales/calibration sessions/recognition peers; membership/assignment-only and public-service profiles; licensed crosswalks; production authority, privacy, retention and native adapters; independent final-file and reference-bridge review.
Subject boundary and candidate types
- PerformanceCycle
- AssessmentMethod
- PerformanceObservation
- Feedback
- Assessment
- Calibration
- Appeal
Deep research questions
- How are observation, contribution attribution, assessment and decision separated?
- Which scales are comparable across roles and periods?
- How are a negative event, the person's response and a revision stored without asserting fault?
Verifiable invariants
- An assessment is bound to a method and a period
- Commit count is not identical to performance
- A disputed claim does not become a confirmed fact by default
End-to-end acceptance scenario
On one case, obtain different justified assessments, preserve an objection and a re-calibration; the public projection does not disclose individual assessments.
Negative case
An automatic employee rating is built from closed tasks without regard to role, complexity and team contribution.
Approaches to compare
- HR Open: exchange of HR data
- W3C ORG and ESCO/SFIA: relationships, skills and levels of responsibility
- HRIS/ATS/LMS practice and independent verification of outcome measurement via SPACE
Candidates in the live catalogue
- WM-ORG-017 · Performance Objective / Review · 0.3.0-research.1 · installable
Semantic fit requires boundary research; a published model does not by itself complete this card.
Result requirements
Every card is executed together with the full research contract: definitions, fields and cardinalities, lifecycle, sources, data mastership, rights, the five object facets, at least eight invariants, positive and negative examples, dependencies, migration and applicability limits.