What an online test can and cannot do
A well-designed online test can give a reasonable estimate of your fluid reasoning. It cannot replace a psychological assessment, which is given one-on-one by a professional, covers several abilities (verbal, memory, speed) and considers context. Only such an assessment allows a diagnosis.
Signs of a good test
- Non-verbal questions (matrices, visual sequences): less dependent on language and culture.
- Progressive difficulty, with genuinely hard questions at the end.
- A time limit: without one, scores are less comparable.
- A normalized score on the mean-100, SD-15 scale, with a percentile.
- Enough questions: ten or so questions cannot measure anything reliable.
- Transparency: the site explains its method and says it is an estimate.
Red flags
- Everyone scores 130 or more (a flattering score to make you pay).
- General-knowledge or vocabulary questions presented as “IQ”.
- A hidden subscription behind a “free test”.
- No mention of the method or the margin of error.
The NexIQ approach
The NexIQ test has 24 non-verbal reasoning questions of increasing difficulty, in 12 minutes. Harder questions count more, the score is placed on the standard scale (mean 100, SD 15) and we always show the margin of the measure. We present it for what it is: a serious estimate, not a diagnosis.
Frequently asked questions
Are free IQ tests reliable?
Some are decent, many are not. Check that they are timed, non-verbal and give a percentile. Be wary of consistently flattering scores.
Can an online test detect giftedness?
It can give a hint, but giftedness is confirmed by a full assessment by a psychologist.
How many questions should a good test have?
At least twenty or so questions of increasing difficulty to give a credible estimate.
Why does my score vary between tests?
Each test has its own margin of error, content and norms. Gaps of 5 to 10 points between two different tests are normal.