Transparent learning design
What we measure, how we guide you, and what we do not claim
Danskertesten organizes preparation around an evidence-based learning loop. Scores are designed to guide the next useful action—not to certify that someone will pass.
Method reviewed: 14 July 2026
The preparation loop
Each step produces evidence for the next one, so preparation is more than a library of questions.
- 1Diagnose a starting point.
- 2Plan a small, realistic week.
- 3Learn the relevant curriculum material.
- 4Retrieve knowledge through questions and spaced repetition.
- 5Simulate the complete exam, including Danish values.
- 6Reflect on errors and evidence.
- 7Re-plan around the clearest gap.
Readiness is an estimate with gates
The estimate has four visible dimensions. High accuracy in one area cannot hide missing evidence elsewhere.
- Knowledge evidence from varied questions
- Coverage across the six exam domains
- Full mock-exam performance
- The separate Danish-values requirement
Sparse evidence caps the displayed estimate at 49; medium-confidence evidence caps it at 79. This prevents a few correct answers from looking conclusive.
A strong-readiness state additionally requires broad coverage, completed mocks, and a demonstrated 4/5 result on Danish values.
How learning evidence is interpreted
We distinguish activity from credible evidence. Objective evidence includes sample size, breadth, and recency.
- Observed accuracy is adjusted conservatively when the sample is small.
- Confidence grows only with attempts across distinct questions.
- Evidence older than 30 days is marked stale.
- Errors link back to the relevant curriculum objective and focused practice.
Known limitations
These limitations are part of the product, not fine print.
- The estimate is not a probability, guarantee, certification, or official exam result.
- Most historical question-to-objective mappings are inferred until human editorial review is complete.
- Current data cannot yet distinguish immediate repetition from delayed retrieval.
- Thresholds must be calibrated against later mocks and consented real-exam outcomes.
- Danskertesten is independent and does not replace official guidance.
Privacy and partner reporting
Learning evidence is learner-scoped. Organization views must use consented, interpretable aggregates—not public rankings or unexplained risk scores. Learners must be able to revoke sharing.
Start with evidence, not a promise
Try a short diagnostic. The first result intentionally stays cautious until there is enough evidence to guide a more specific plan.
Try the diagnostic