IQ Test for Scientists

Scientists and researchers can take this IQ test in about 12 minutes: 25 abstract reasoning puzzles, no sign-up to start.

What's your IQ?

Take the IQ test

25 questions · about 12 minutes

What would you score? Try question 1.

Which option completes the grid? One tap, no sign-up.

The test at a glance

Questions: 25 visual puzzles
Question types: 5 sets, each easy to hard
Time: 12-minute timer
Scoring: Same for every age, not age-adjusted
Cost: No sign-up to start. When you finish, you choose a report to see your result.

Scores are not age-normed, and this is not a clinical or diagnostic test.

Scientists are often the people most sceptical of IQ tests, which is healthy.

Here is what the research on scientists and measured intelligence actually shows, and how to read a result from a short test like this one.

Two Nobel laureates who did not make the cut

In the 1920s Lewis Terman at Stanford selected more than 1,500 California children with very high IQ scores and followed them for decades. Two boys who were tested but scored too low to be included, William Shockley and Luis Alvarez, both went on to win the Nobel Prize in Physics. None of the children selected did.

In 1953 the psychologist Anne Roe published a study of 64 eminent American scientists. Their test scores were high but varied a lot, and what they shared most was intense absorption in their work.

How this test works for scientists

This is a short test of fluid reasoning with 25 items. You will want to know its limits: it cannot match the reliability of a full clinical battery, its score is not age-adjusted, and it measures visual reasoning only, not verbal or quantitative ability.

Treat the result as a single noisy measurement.
The items are matrix-style and progressive: each set of five rises in difficulty.

What scores mean for scientists

Researchers tend to score above the population average, but there is no reliable modern figure for an "average scientist IQ".

The Terman and Roe findings point the same way: above a certain level, persistence, curiosity and opportunity explain more about who makes discoveries than extra IQ points do.

Reading your result like a measurement

Every IQ score comes with measurement error. For a full clinical test such as the WAIS-IV, the standard error of the full-scale score is only around 2 to 3 points, because it combines many subtests. A 25-item test has far fewer data points, so its error is wider, and it is widest at the extremes.

Practice effects add a bias on top: a second attempt usually scores a few points higher. The scientific way to read a single result is as a range, with a small upward bias on any retake, not as a precise value.