Wonderlic Score to IQ: What Converts and What Does Not

Ask what a Wonderlic score of 30 is "in IQ" and the internet answers instantly: 120. The arithmetic is a raw score doubled plus sixty, it is quoted in draft coverage every year, and it is repeated confidently enough that most people assume it came off the back of the test booklet. It did not. Understanding why it is wrong is a short lesson in what a converted score can and cannot carry.

What the test actually is

The Wonderlic is a brief cognitive ability test built for employment screening: 50 items in 12 minutes, mixing verbal, numerical and spatial questions of rising difficulty. The result is a raw count of correct answers, from 0 to 50. Very few people finish. That is deliberate — the time limit is a working part of the instrument, not an administrative convenience.

Note what is missing. There is no mean of 100 and no standard deviation of 15. A raw count is not on an IQ scale at all, so there is nothing to report as an IQ without first mapping the raw distribution onto one. That mapping requires a norm sample and a published conversion table, and the question is whether the popular formula is one.

Where the formula comes from, and what it does

It comes from repetition. The double-and-add-sixty rule spread through American football draft coverage, where Wonderlic results were widely reported and readers wanted them in a familiar unit. It is not a published conversion, it is not derived from a linked norming study of both instruments, and the publisher's own materials report results as raw scores and percentiles against occupational norm groups.

The rule survives because it is roughly calibrated in the middle. A raw score in the low twenties is around average for many working populations, and the formula returns something near 100 there, which feels right. Its behaviour at the edges is the tell:

• A perfect 50 converts to 160, a figure four standard deviations above the mean on an SD-15 scale — a region containing a few people in a hundred thousand. A short timed test with a ceiling cannot resolve anything at that level.
• A raw score of 10 converts to 80, which the formula places well below average, when a score that low on a strictly timed test may reflect pace or unfamiliarity as much as reasoning.
• The rule is perfectly linear, whereas raw scores on a test with a ceiling and a floor are not linearly related to a normal scale anywhere near the ends.
• It carries no error term. A converted figure quoted to the point inherits the uncertainty of the original and then hides it — the problem described in our note on why a score is a range rather than a number.

Speed is doing more work than people think

Twelve minutes for fifty items is about fourteen seconds each. At that pace the test is measuring how quickly someone can recognise a problem type and execute it, which is a real and job-relevant capacity — but it is not the same construct that an untimed or generously timed reasoning battery measures. Processing speed and reasoning are related, not interchangeable, which is the same distinction our piece on working memory and reasoning draws for a different pair of abilities.

The practical consequence: two candidates with genuinely similar reasoning ability can produce Wonderlic raw scores several points apart because one works faster under pressure. Convert both with a linear rule and the gap is presented as a difference in IQ, on a scale where it implies something it does not.

What you can legitimately compare

Percentiles. A percentile states a position in a named reference sample, and it is the one currency every properly normed test can be converted into without inventing a relationship between them. A Wonderlic raw score sits at some percentile of its norm group; a Wechsler-type score sits at some percentile of its standardisation sample. Those are comparable statements, with two caveats that have to travel with them.

• The norm groups differ. Employment norms are drawn from applicant or incumbent populations, which are not the general adult population, so the same percentile means something different.
• The constructs differ. A 12-minute speeded screen and a two-hour individually administered battery do not measure the same mixture of abilities, however well their percentiles line up.
• Both figures carry measurement error, and a short test carries more of it.
• A percentile is not a probability of success at anything. It is a position, and the link from position to outcome is a group average with wide spread around it.

If you want a defensible number from a score you hold, convert it the honest way: state the scale, find the percentile against a named norm group, and keep the interval. Our IQ score converter handles translations between scales that genuinely share a normal metric — standard deviation 15 to 16, or to the Cattell scale — which is exactly the class of conversion the Wonderlic rule is imitating without having earned.

The broader rule is worth carrying out of this. A score converted between instruments is only as good as the study that linked them, and if nobody can name that study, the conversion is folklore with a decimal point. That applies to the Wonderlic rule, to the figures attached to famous names, and to any number that arrives already translated into a unit its own test never used.

Common questions