An open mind.
An honest method.
We make original reasoning tasks for personal exploration. A useful experience starts with a clear account of what the result can, and cannot, tell you.
What the assessment contains
A session selects 24 tasks across abstract reasoning, matrix reasoning, number sequences, pattern recognition, spatial reasoning, and logical deduction. Four tasks per category provide a small, descriptive sample. Tasks are balanced across three construction difficulty levels. Those levels describe rule complexity; they have not been calibrated against a population.
How tasks are created
Our initial bank uses 11 original rule families: rotation, reflection, shape progression, counting, color transformation, position transformation, combination, XOR, symmetry, multiple rules, and numerical transformation. A seed reproduces the item, its choices, and its construction rule. Automatic checks look for duplicate choices, inconsistent rule solutions, invalid structures, and repeated items. Automated checks do not replace independent human review. Generated items are marked as uncalibrated. No protected Raven, WAIS, or Mensa items are included. External open banks are not imported unless the exact materials, attribution requirements, and commercial use rights have been reviewed.
How your result is calculated
Our server checks your submitted choices against the frozen version of each item. Each correct choice contributes one point. Skipped and incorrect items contribute zero points. We record total correct answers, observed category accuracy, recorded response times, and the scoring version. Response time does not alter the score.
What we do not report
There is currently no representative normative sample, calibrated item response model, or validated conversion to IQ. We therefore do not report a standardized IQ score, population percentile, mental age, clinical confidence interval, or diagnostic classification. The full report adds detail to observed performance; payment does not make the assessment scientifically validated.
How to read category differences
Four items per category cannot establish a stable strength or weakness. A one-item difference changes category accuracy by 25 percentage points. Task familiarity, chance, attention, device, language, accessibility needs, and testing conditions can affect results. A stronger thread is identified only when sufficient responses are present and there is no tied highest category.
What the Cognitive Signature means
The signature is a visual interpretation of your six observed category accuracies. Its geometry is a creative way to revisit your session. It is not a biometric identifier, personality model, brain scan, or independently validated psychological profile.
Before standardized interpretation
Future standardized scores require expert review, a defined intended use, representative recruitment, sample size planning, reliability and validity studies, item calibration, differential item functioning analysis, and documented norm limitations. Adaptive selection and IQ reporting remain disabled until their prerequisites are met. We will version any change and retain the method used for each result.
Appropriate use
This product is for adults 18+ and personal exploration. Do not use it for diagnosis, hiring, education placement, immigration, legal decisions, or other consequential judgments. For a formal evaluation, consult an appropriately qualified professional using validated instruments.