Many people think IQ tests do not measure creativity. This belief is ironic because IQ tests are backed by the statistical concept known as g (the hypothesized general factor that causes all mental abilities to positively correlate) and the most g loaded tests are those that require you to see associations between seemingly unrelated things: Pattern recognition.

Even g loaded tests as tedious as vocabulary or general knowledge require creativity because acquiring a large fund of information requires you to make some creative associations. For example, if someone asks to borrow some “dead presidents”, you must associate borrowing with money and then be creative enough to associate money with the pictures on currency which are of historical presidents. Hence “dead presidents” enters your vocabulary as a synonym for money.

But because IQ tests all have the same right answer that all high IQ people by definition converge on, many feel that they can’t be measuring creativity (which implies original thought). As a result, psychologists have created divergent thinking tests which supposedly measure creativity. A typical divergent thinking test is to ask people to name as many uses for a brick that they can think of in two minutes. Such tests do not have one right answer or even 100 right answers. The number of right answers is potentially infinite. Original answers like “for a short man to stand on when kissing a tall girl”, or “to put in your suitcase when you leave a hotel without paying so they think you’re still there” get more credit than unoriginal answers like “smash a window”, or “help build a house”.

Although divergent thinking tests correlate positively with conventional IQ tests, the correlation is low (and some say it vanishes altogether above IQ 120).

But one reason for the low correlation could be that divergent thinking tests are not measuring a cognitive ability but rather a personality trait. According to Arthur Jensen, in order for a test to be measuring an “ability” (physical or mental), there must be a clear standard of proficiency. Everyone can agree that remembering five digits is more impressive than remembering two digits or that solving a puzzle in 2 minutes is more impressive than solving it in 3 minutes. But can everyone agree that using a brick to kiss a tall girl is more impressive than using a brick to build?

There are humor tests where there is no single right answer. For example people are asked to write a caption to a cartoon and the funnier the caption, the higher the score. This potentially does have a clear standard of proficiency because although humor is subjective, laughter is involuntary, and if the test participants knew the objective was to make as many people laugh as possible, this might make a good psychometric measure of creativity.

Of course it would be completely impractical because every time someone was tested, you’d need to poll a representative sample of the public on whether the person’s answers were funny. But given that stand-up comedians have high IQs, I highly suspect this test would correlate at least moderately with g.