Assessment Criteria
The assessment measures professional judgment, not how professional the drawing looks. Five criteria across four assessment levels. Verification against primary sources and professional accountability carry the most weight.
| Criterion | Outstanding | Good | Developing | Insufficient | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 1. The human anchor (solution from own trade, without machine) | Clear, professional, identifies real hazards; utilized in comparison | Present and utilized | Superficial | Missing or written after the fact | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| 2. Verification against regulation | Every numerical reference confirmed or refuted with a named primary source | Most numbers verified | Inconsistent, some numbers unconfirmed | Machine numbers accepted as truth | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| 3. Icelandic reality | Clearly points out false assumptions (frost depth, materials, load) | Analyzes at least one | Mentions without analysis | Accepts foreign assumptions as valid | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| 4. Honest Oracle (resistance) | Identifies hallucinated standards/numbers, rejects ≥2 points with reasoning | Rejects with reasoning | Rejection is shallow | No criticism of the proposal | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| 5. Professional accountability | Clear statement of accountability: what holds up, what does not, who is responsible | Accountability present | Generically phrased | Missing — machine is "blamed" | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| (Fill-in template table) | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Overall Grade (Guideline)
- Excellent: 4–5 criteria outstanding, nothing below "Good".
- Good: most "Good" or better, verification and accountability present.
- Pass: methodological locks present but shallow.
- Fail: machine numbers accepted as truth, no verification or no accountability.
Key rule: a number without a verified primary source cannot be considered "correct". And an empty statement of accountability cannot score higher than "Developing" — accountability is the core of the trade.