That is, until psychologist Joseph Fagan published a paper in the 1980s arguing that traditional IQ tests, the kind developed by Binet and modified by David Wechsler, created the illusion of coherence where there wasn't really any. That is, traditional IQ tests measure concrete knowledge ("What is the capital of Pennsylvania?") with more basic processing skills (remembering a list of words). Researchers have found again and again that while each person taking the test might do better on some parts of the test than others (For instance, I always do terribly at creating a visual pattern to match a picture, but I do well at analogies), there is, generally speaking, a lot of consistency— in other words, a high correlation between components of the test. The person who gets a higher score than others her age on one part is likely to get a higher score than others on most of the other parts of the test. It is easy to see how scientists, and ultimately the general public, came to think of this test as actually measuring a particular quality of mind or even a physical part of the brain.
Psychologists have even given this imagined underlying quality a name: g (for "general intelligence"). If you're high in g, you are smart, and if you are high in g, you are likely to do well on many components of the test. Fagan didn't disagree that intelligence might ultimately be a single quality of mind, but he wanted a test that would actually focus on just that quality, the ability that produces g. So he zeroed in on the single characteristic he thought underlay the myriad of abilities we push together and call intelligence.
Fagan argued that what makes one person do better on an IQ test and seem smarter in real life as well is what he called speed of processing. We are all familiar with that concept from our computers: the faster the processor, the more the computer can do. It's the same with the human brain. The faster it can take in information, the more information it can take in. Hence two children might be exposed to the same environment, but the one who can take in more will know more. Fagan's point was that speed of processing is a much simpler, more precise, and more value-free characteristic, which might actually explain the correlation between items on traditional IQ scores. But how do you directly measure something like speed of processing?
This is pretty easy, as it turns out. From birth, babies stare at something until they become used to it—in other words, until they have processed it. Then they look for new stimuli. Fagan showed babies two pictures projected onto a screen in front of them. Then he measured how long it took them to absorb (become familiar with, or process) the first picture before turning their heads to look at the second picture. Of course, it's possible that some children simply have shorter attention spans than others. And yet the important thing about Fagan's test was that there was enormous consistency between his infant test and more traditional IQ tests. Babies who processed visual information quickly on Fagan's test also did well on IQ tests when they were in elementary school.