Transparency
Methodology and limitations.
Atlas Intelligence is designed for personal reflection on performance within one versioned reasoning assessment.
Assessment structure
The assessment contains 28 roles across Abstract, Logical, Numerical, and Spatial Reasoning. Each domain contains six scored items and one zero-score calibration item. Exactly one candidate is assigned per role.
Deterministic scoring
Each domain has a raw score from 0–6. Display values are fixed: 0 → 0, 1 → 17, 2 → 33, 3 → 50, 4 → 67, 5 → 83, and 6 → 100. Calibration responses and response time do not affect scores or report interpretation.
What the result means
The report compares performance across the four domains within this assessment. It may describe relatively stronger results, lower results, ties, and balanced or varied profile shapes when supported by the scores.
Important limitations
- It is not an IQ score, percentile, or population rank.
- It is not population normed and does not estimate a precise psychometric trait.
- It is not diagnostic and is not intended for medical, hiring, placement, or educational suitability decisions.
- The Human Pilot has not started and the Production Question Bank is not approved.
- Results can reflect the assessment content, conditions, attention, and ordinary measurement uncertainty.