Bloomberg reports on an intriguing device that’s apparently being used by an increasing number of doctors to weed out college students faking ADHD in order to score prescriptions to stimulant medications. It’s called the Quotient ADHD System, and it collects data on a person’s ability to “sit still, inhibit impulsivity and respond accurately to images on a computer screen.” Developed at McLean Hospital, the psychiatric wing of Harvard Medical School, it uses motion detection technology, along with the patient’s performance on a visual response test, to measure inattention, hyperactivity and impulsivity compared to other people of the same age and gender.

Some doctors swear by it—here’s one who says it proved him wrong on several cases and here’s a Time reporter who tried it—and the manufacturer claims that it’s 92 percent accurate in identifying fakers. But it hasn’t exactly taken the ADHD world by storm. After about 5 years on the market only about 300 are in use, according to Bloomberg.

Quotient wasn’t developed, of course, to weed out students trying to score Ritalin or Adderall to boost their academic performance, but rather to improve diagnoses for the millions of children who have real problems. The goal was a test that would be less laborious and subjective than the scales now used to have parents, teachers, and children themselves measure how often they exhibit the behaviors that are symptoms of ADHD.

Those tests—like the SNAP-IV Teacher and Parent Rating Scale and the Child Behavior Checklist—are limited by the fact that answers can be skewed by preconceptions. That’s why the Quotient people like to taut their test as “objective.” But the bigger limitation in the use of scales is that too many doctors don’t take the time to administer them, and hence too many kids are diagnosed basically by hunch—and on the hope that medication will help them do better. And when it comes to college kids, clinicians surely aren’t asking for the multiple sources of information that are considered best practices when diagnosing kids.

Quotient has the appeal that it takes only about 20 minutes, and if it works to weed out the real inattention and hyperactivity from the fake, it could be useful for cutting down on abuse of the meds. But as Dr. Rachel G. Klein, professor of child and adolescent psychiatry at NYU’s Langone Medical Center and a member of our Scientific Research Council, told Bloomberg, to be diagnosed with ADHD a child must show symptoms over a period of time and in several settings—not just at school, for instance, or at home. That fact that a child has symptoms of ADHD on a visit to a doctor’s office isn’t the basis for a sound diagnosis. We’re all for developing objective tests, but the fact is that the science isn’t there yet for us to make a diagnosis with a machine. Until we can, doctors should use the best evidence-based tools and the secret weapon of every good clinician: rigor.