Method before marketing

How the IQTestHQ estimate works

IQTestHQ is a focused online reasoning test. It is designed to be useful and transparent, not to imitate the authority of a supervised clinical assessment.

What the test measures

The first version uses 36 original questions, nine in each of four domains: spatial visualization, pattern reasoning, quantitative and sequential reasoning, and verbal and deductive reasoning. These tasks sample important reasoning skills, but they do not cover creativity, practical judgment, memory, knowledge, motivation or every theory of intelligence.

Question design and copyright

The format families are informed by published cognitive-test practice, including rotation, matrix, sequence and verbal-relation tasks. All live prompts, answer options, graphics and explanations are authored for IQTestHQ. We do not use ICAR item content, Raven's Progressive Matrices artwork or copied questions from other online tests.

Version 1 scoring

Each correct answer counts equally. Domain subscores are the number correct within that domain. A fixed, monotonic and versioned lookup converts the 36-item raw score into an IQ-style estimate and percentile. Expert difficulty labels help balance the form, but are not used as unmeasured score weights.

Important: the launch score model is provisional. It is not yet based on a representative population norm. Read the single score as a range with expected uncertainty of approximately ±10 points.

Fairness and known limits

  • English verbal items are affected by language and education.
  • Screen size, accessibility needs, fatigue and interruptions can affect performance.
  • Repeated attempts can benefit from practice and are not directly comparable to a first attempt.
  • Online, unsupervised conditions cannot verify identity, assistance or test environment.

Calibration plan

Only eligible first attempts enter the calibration dataset. After at least 5,000 valid first attempts, we will review item difficulty, discrimination, reliability, timing anomalies and model fit before changing any weights or mapping. Issued reports keep their original test and scoring version; model changes are never applied silently to old results.

What the result is not for

Do not use IQTestHQ to diagnose a condition, make a medical decision, qualify for Mensa, select a student or employee, or replace a supervised assessment by a qualified professional.

Ready to try it? Start with the untimed practice question.