About the test
Matrix reasoning test
Build the missing figure in visual matrices and see how many tasks you solve. The tasks come from the Open Matrices Item Bank (OMIB), an open research item bank published in 2022. Our quick, medium and long forms are pilots, so you get the number of correct answers, not an IQ score.
Questionnaire / instrument: OMIB
The test at a glance
- Price
- Free, including your results
- Registration
- No account required
- Length
- 22 tasks in the standard form, or 12 or 36
- Time
- About 20 minutes, with no time limit
- Format
- Build the answer from 20 elements · two practice tasks
- Your report
- Correct answers · answered and skipped tasks
Three test lengths
Every task is a three-by-three matrix with the last square missing. The rules include adding, combining, subtracting and rotating elements, or completing a pattern. You choose the length, and the report shows which form you took:
Quick: 12 tasks
Twelve tasks from the authors’ set 2, chosen to span its published difficulty range. About 8–12 minutes. A first impression, but with few tasks a single slip matters more.
Medium: 22 tasks
The recommended form: all 22 tasks of the authors’ original set 2. About 15–25 minutes, including practice.
Long: 36 tasks
36 tasks chosen from the whole item bank to span its published difficulty. About 25–40 minutes. More tasks give a broader sample of patterns, but this form is also a pilot.
What your results show
The report describes your performance on the form you took, with notes on how to read it. You can preview the layout before you start.
- The number of correct answers out of the number of tasks, for example 14 of 22.
- How many tasks you answered and how many you skipped. Skipped tasks count as incorrect.
- The test form and version you took.
- An explanation of why the score is not converted into an IQ and what can affect it.
What does it measure?
Non-verbal reasoning with visual matrices. You identify the rules in a pattern and construct the missing element. This samples one aspect of cognitive ability rather than the full range measured by a comprehensive intelligence assessment.
Where does it come from?
The Open Matrices Item Bank (OMIB) was published by Koch, Spinath, Greiff and Becker in 2022. We use its open item bank and have adapted the presentation for this website.
How is the result calculated?
Two practice tasks with feedback come first and are not scored. In the main test you build each answer from 20 elements, and a task earns one point only when your selection matches the key exactly. Skipped tasks count as zero. There is no time limit and no feedback during the main test. The default 22-task form is the authors’ original set 2; the 12-task form selects tasks from set 2 spread across their published difficulty, and the 36-task form selects tasks spread across the whole bank. Your score is the number of correct answers.
What can the result tell you?
These forms are pilots, not three separately validated tests. We report correct answers, not a normed IQ score or population percentile. Published item parameters are not used to estimate ability in the current scoring.
Languages and adaptations
The matrix tasks use visual elements rather than words. Instructions and explanations are translated for this website. A non-verbal format does not make performance independent of culture, prior practice or testing conditions.
Frequently asked questions
Is this an IQ test?
It uses the kind of visual matrix tasks found in many intelligence tests, but it does not give an IQ. It samples one aspect of reasoning: finding rules in unfamiliar visual problems. A full intelligence assessment covers several abilities and is carried out under controlled conditions by a qualified professional.
Why do I not get an IQ score?
An IQ score places you relative to a representative sample of the population. The original study does not provide general-population IQ norms, and we have not built norms for our forms. The authors published item difficulty parameters, but we do not use them to estimate ability yet, because the forms and the estimation have not been checked independently.
Is the test free?
Yes. You can take any of the three forms and see your result without paying or creating an account.
Can I compare scores from the quick, medium and long forms?
No. The forms have different numbers of tasks and different selections of difficulty, so 9 of 12 and 15 of 22 cannot be compared directly. Compare a result only with another attempt on the same form, and keep in mind that you will usually do better the second time through practice.
How is this different from the logical reasoning test?
The logical reasoning test (SAFRAN) has 54 verbal, figural and numerical problems with a 15-minute limit and multiple-choice or number answers. This test has only visual matrices, no time limit, and you build each answer yourself from 20 elements. Both report correct answers, not IQ.
What affects my score?
Concentration, interruptions, tiredness, the device you use and earlier experience with similar puzzles can all affect the result. Work on your own without help, and choose a time when you can finish without being disturbed.
Why are you asking for my age, gender and country?
These questions are optional and do not change your score. If you answer them, they help us evaluate who the test works for and build a better basis for comparisons later. They are not shown on shared result pages.
License and reuse
GNU GPL v3The OMIB materials and our adapted asset generator are distributed under GNU GPL v3. Redistribution must preserve the applicable GPL terms and source availability. The research article’s CC BY license is separate from the license for the test materials.
The terms below apply to the test materials. Articles, translations, software and our result descriptions may have separate terms.
Sources and further reading
Koch, M., Spinath, F. M., Greiff, S., & Becker, N. (2022). Development and Validation of the Open Matrices Item Bank. Journal of Intelligence, 10(3), 41. https://doi.org/10.3390/jintelligence10030041