Variety of IQ Tests Means Measuring Gray Matter Is a Gray Area, By JO CRAVEN MCGINTY
Measuring smarts is harder than it sounds.
IQ tests are widely accepted measurements of intelligence, but there are many different tests, their content varies, and a person’s scores may fluctuate for a variety of reasons. In most cases, the differences won’t be extreme. But researchers have documented scores for individuals that varied by 10 points in either direction.
“It’s sloppy, and we know that,” said W. Joel Schneider, a psychologist at Illinois State University who blogs about intelligence tests at Assessing Psyche. “It’s too big of a concept to measure precisely in a way we can all agree. Think of a cheap bathroom scale. Every time someone steps on it, it’s slightly different.”
Still, the tests are used to screen children for gifted or special education programs, to determine whether prospective employees are right for a job, and to decide whether someone convicted of a capital offense can be executed.
IQ tests are actually a collection of assessments intended to measure different abilities, such as logic, pattern recognition, verbal aptitude, spatial orientation and short-term memory. But multiple companies publish different tests, and they don’t all measure the same thing. Some emphasize certain abilities over others. Some measure fewer abilities than others. And the narrow focus of abbreviated tests often used for screening could make someone look more, or less, gifted than a full test.
“A lot of people think there is this one test called The IQ Test,” Dr. Schneider said, “but there are many tests.”
Unlike academic achievement tests, which are meant to evaluate what students have learned in school, IQ tests are intended to measure general cognitive ability. A full-scale test could assess seven different abilities or twice that many—there are no hard and fast rules. An abbreviated test might examine only a couple of abilities.
“A lot of people involved in gifted education run into problems with that,” said Kevin McGrew, the Director of the Institute for Applied Psychometrics and a co-author of the Woodcock-Johnson Battery III and IV IQ tests. “You may have a child with an exceptional ability in a certain area, but it’s not measured by the screening test.”
Modern intelligence tests were introduced in France at the turn of the last century to identify children who were likely to struggle in school. The performance of an individual was assessed by calculating the difference between mental and chronological age. An 8-year-old whose score was typical of a 6-year-old, for example, was said to have a mental age that was two years behind.
But mental age levels off after a certain point while chronological age steadily increases throughout life, and psychologists soon realized that expressing the performance as a ratio would be more useful.
Within a decade or so, the convention changed. Mental age was divided by chronological age and then, to get rid of the decimal place, multiplied by 100. The result was by definition a quotient, and the score became known as the intelligence quotient, or IQ.
By this measure, an 8-year-old who performed only as well as a 6-year-old would have an IQ of 75, while a 14-year-old performing at the level of a 12-year-old would have an IQ of 86. This was an improvement, but the scoring still didn’t work for adults. It doesn’t make sense, for example, to relate the mental age of a 40-year-old to that of a 38-year-old.
Today, the term “IQ” is a relic. Individuals are measured against people of the same age on a scale that assumes IQs are normally distributed with an average score of 100 and a standard deviation of 15. (Some tests use a different standard deviation.) Scores are reported with a margin of error. In other words, they are an estimate.
In a normal distribution, two-thirds of individuals will score between 85 and 115. About 95% will fall between 70 and 130. And about 2.5% will fall below 70, the traditional threshold for intellectual disability, or above 130.
Nowadays, IQ tests may be used to assess whether potential employees are right for a job, whether someone qualifies for some Social Security disability benefits or whether a person convicted of a capital offense can be executed. National Football League players, for example, are routinely administered the Wonderlic test, which is intended to predict workplace success. And in 2002, the U.S. Supreme Court ruled that it is unconstitutional to execute people who are intellectually disabled.
Student performance on the best available IQ tests is correlated with academic success, Dr. McGrew said, but he noted that even the best tests explain only about 40% to 50% of school achievement.
“That’s a lot in psychology,” he said. “But it means 50% to 60% is not related to cognitive ability.”
Although no one discounts the importance of raw intelligence, it turns out that qualities such as motivation, determination and a desire to succeed—qualities that IQ tests don’t measure—play a significant role in success.
This isn’t exactly new information. Thomas Edison once summed it up like this: Genius, he said, is 1% inspiration and 99% perspiration.
Write to Jo Craven McGinty at Jo.McGinty@wsj.com