How the score is decided

36 non-verbal items are arranged in six blocks of rising difficulty, drawing on four families of reasoning:

• Pattern completion — Reading two rules at once across rows and columns.
• Sequence logic — Spotting how a property advances step by step.
• Spatial rotation — Tracking an object as it turns through space.
• Figure logic — Combining or cancelling the parts of two figures.

Every item is drawn from a rule rather than stored as a picture, so each sitting gets freshly ordered alternatives. Nothing is reproduced from any published test.

Counting every item equally would treat the first warm-up puzzle as worth the same as the hardest one on the form. Instead each item carries a designed difficulty weight from 1 to 5, and your raw score is the sum of the weights of the items you answered correctly. Skipped items score zero.

The raw score is converted to a z-score — how many standard deviations you sit from the expected middle of the distribution — and then placed on the conventional scale.

The norm reference currently assumes an average taker clears about 42% of the weighted points, with a spread of roughly 18% of the total. That figure comes from the designed difficulty of the form, and will be recalculated from real attempts as they accumulate. Scores are capped to the 55–160 range because no reference sample supports claims beyond it.

A score of 100 is the middle of the distribution. Each 15 points is one standard deviation — so 115 places you above roughly 84% of people, and 130 above roughly 98%.

A professional assessment is administered one-to-one by a licensed psychologist over one to two hours and scored against a demographically stratified sample. MentalClarity is self-administered, unsupervised, and standardised against a designed reference rather than a validated norm group.

Treat the number as an informed estimate of non-verbal reasoning. It is not a diagnosis, and it should not be used for medical, educational, employment or legal decisions.