IQ by itself is not a very good predictor of job performance. Hunter and Hunter (1984) estimated that at best IQ accounts for about 25 percent of the variance. Sternberg (1996) has pointed out that studies vary and that 10 percent may be a more realistic estimate. In some studies, IQ accounts for as little as 4 percent of the variance.

An example of this research on the limits of IQ as a predictor is the Sommerville study, a 40 year longitudinal investigation of 450 boys who grew up in Sommerville, Massachusetts. Two-thirds of the boys were from welfare families and one-third had IQ.s below 90. However IQ had little relation to how well they did at work or the rest of their lives. What made the biggest difference was childhood abilities such as being able to handle frustration, control emotions and get along with other people.

Another good example is a study of 80 Ph.D's in science who underwent a battery of personality tests, IQ tests, and interviews in the 1950s when they were graduate students at Berkeley. Forty years later, when they were in their early seventies, they were tracked down and estimates were made of their success based on resumes, evaluations by experts in their own fields, and sources like American Men and Women of Science. It turned out that social and emotional abilities were four times more important than IQ in determining professional success and prestige.

Now it would be absurd to suggest that cognitive ability is irrelevant for success in science. One needs a relatively high level of such ability merely to get admitted to a graduate science program at a school like Berkeley. Once you are admitted, however, what matters in terms of how you do compared to your peers has less to do with IQ differences and more to do with social and emotional factors. To put it another way, if you're a scientist, you probably needed an IQ of 120 or so simply to get a doctorate and a job. But then it is more important to be able to persist in the face of difficulty and to get along well with colleagues and subordinates than it is to have an extra 10 or 15 points of IQ. The same is true in many other occupations.

We also should keep in mind that cognitive and non-cognitive abilities are very much related. In fact, there is research suggesting that emotional and social skills actually help improve cognitive functioning. For instance, in the famous marshmallow studies at Stanford University, four year olds were asked to stay in a room alone with a marshmallow and wait for a researcher to return. They were told that if they could wait until the researcher came back before eating the marshmallow, they could have two. Ten years later the researchers tracked down the kids who participated in the study. They found that the kids who were able to resist temptation had a total SAT score that was 210 points higher than those kids who were unable to wait.

Granted that cognitive ability seems to play a rather limited role in accounting. In doing the research for his first book, Goleman (1995) became familiar with a wealth of research pointing to the importance of social and emotional abilities for personal success. Some of this research came from personality and social psychology and some came from the burgeoning field of neuropsychology.