Skip to content
FreeIQTest.blog
Take the free test

How is the FreeIQTest IQ test built and scored?

Written by FreeIQTest · Reviewed by Not yet reviewed · Updated

Short answer

The FreeIQTest IQ test has 48 original reasoning puzzles generated from rules. Correct answers earn more points on harder questions; the total is converted to the IQ scale (average 100) with provisional norms and shown with a ±10-point range. It is an estimate, not a clinical assessment, and the norms will be recalibrated on real results.

What does the test contain?

The test contains 48 puzzles in four types, mixed together and ordered from easier to harder. You can take the free IQ test to see each type.

Puzzle types in item set 2026-11-v1
Puzzle typeWhat the rules varyItems
Pattern matricesShape, number of shapes, fill and rotation across rows and columns14
Number seriesAdding, multiplying, growing gaps, squares, cubes, alternating series12
Rotated shapesRotation versus mirror image; harder items add near-miss shapes10
Logic puzzlesSymbol equations and odd-one-out figures12

How are the questions made?

Every question is produced by a program from explicit rules, so no question is copied from another test or website. Each item gets a design difficulty from 1 to 5 based on how many rules it combines; for example, a matrix where only the shape changes from row to row is level 1, and one where shape, fill and number all follow different patterns is level 5. These levels are our estimate until real response data shows how hard each item actually is.

How is the score calculated?

The score is calculated in five steps:

  1. Each correct answer earns points by design difficulty: 0.75 for level 1, rising by 0.25 per level to 1.75 for level 5. Wrong and skipped answers earn 0, and there is no penalty for guessing.
  2. Your points are divided by the maximum possible points to give a proportion between 0 and 1.
  3. That proportion is converted to the IQ scale with a provisional norm (explained below).
  4. Your percentile is the share of a normal distribution with mean 100 and standard deviation 15 that falls below your score.
  5. The likely range is your score ±10 points.

Scores are limited to the range 55–145, because a 48-item test can't separate scores reliably beyond it.

What are provisional norms?

A norm is the reference that turns a raw result into an IQ score by comparing it with how other people did. We don't have that reference yet, so the test uses a stated assumption: a typical adult earns half of the available points, and one standard deviation (15 IQ points) equals 18% of the available points.

This assumption is a placeholder, not a measurement. Once enough anonymous results are in, we will estimate the real distribution of scores, recalibrate the conversion and item difficulties, publish the sample size and date range here, and label results accordingly.

How should you read your result?

Read your result as a rough estimate within its ±10-point range. An unsupervised online test can't control for distractions, guessing, devices or practice, and taking the same questions again usually raises the score. For a diagnosis, giftedness assessment or school placement, see a licensed psychologist, who will use a standardised, individually administered test.

What happens to your answers?

Your answers are scored on our server, and the correct answers are never sent to your browser. If result storage is switched on, we keep an anonymous record of the session (language, country, device type, timing) and the result, with no name, email or IP address. The privacy policy lists exactly what is stored.

What will change next?

We will publish a change log on this page whenever the items or the scoring change. Planned changes are norms estimated from real results, item difficulty estimates from response data, and a measured reliability figure to replace the fixed ±10 range.