A Revolution in Psychological Testing

Money in Washington is tight, and researchers who compete for federal grants are feeling the pinch. The National Institute of Mental Health (NIMH), for example, now funds only 6 to 8 percent of the applications it receives, and typically only those that relate to severe mental illness, according to Robert Bornstein, Ph.D., a professor at Adelphi’s Gordon F. Derner Institute of Advanced Psychological Studies.

So how did Dr. Bornstein and his colleague Gregory Haggerty, M.A. ’03, Ph.D. ’08, assistant director of residency training at Nassau University Medical Center, win a two-year, $320,000 grant from the agency to evaluate tests of narcissism and dependency? “Someone there looked at [our proposal] and said, this is transformative,” Dr. Bornstein said.

The paradigm shift, as Dr. Bornstein described it, has to do with how psychiatric and psychological tests are validated. For the past 60 years, experts have used the same flawed method. They introduce a new test—for intelligence or social phobia or narcissism—and validate it by seeing how well it correlates with existing measurements for the same thing. A new intelligence test, for example, might be compared to SAT scores and GPAs.

The problem, Dr. Bornstein explained, is that new tests always correlate with a broad array of other traits. A new test for narcissism, for example, will correlate with self-esteem and self-confidence, making it hard to tease these qualities apart.

“This is where we had this lightbulb moment,” Dr. Bornstein said. “What we need to do is come up with an experimental manipulation that primes the construct that you are trying to measure.”

By using subliminal methods, researchers can exaggerate certain qualities in their subjects. For example, hearing narcissism-related words will boost someone’s narcissism. Dependency-related words will temporarily elevate someone’s sense of dependency, and so on.

Dr. Bornstein explained: “If you have a 48-item measure of narcissism, you administer it to a bunch of patients, wait two weeks, call them back, prime narcissism, administer it again. Some of those 48 items are going to move up in response to the prime. Others are not. Those ones that didn’t were noise. They were error variance in your measure. So now we can turn your 48-item measure into a 22-item measure that does better.”

Once the method has been tested and proven, the National Institutes of Health (NIH) and others can adapt it for any psychological or psychiatric measure. “Once the method becomes routine, you can easily deliver it to any researcher who wants to make a large and kind of messy measure less messy,” Dr. Bornstein said. In the long run that new method can have effects more far-reaching than the study of any individual syndrome or disorder.