Think you're mentally normal? Think again

By: Sam Ohmer

Posted: 5/1/08

It's fairly normal for people to wonder whether they are "normal" or not. When performing diagnostic tests or psychological research studies, it is important for doctors to be able to define and determine "normality" as accurately as possible.

A battery of tests - from brain imaging studies to preference surveys to diagnostic tests for mental illness - are dependent on the definition of normal and abnormal scores or outcomes. Usually these values are determined by experts after many rounds of testing.

The problem is, for years researchers have noticed that supposedly normal people don't always test normally. Even the epitome of healthy brain activity can be a little tilted, according to test results.

That poses a challenge for doctors and researchers alike: How are we to divine true abnormalities when even relatively healthy people are giving whacky test results? A new study from Barry Gordon, a Hopkins neurologist, and colleagues at Yale University and Hartford Hospital, both in Connecticut, offers some insights into this phenomenon.

First, they found that the mere fact that someone tests abnormally does not mean they are brain damaged. Healthy people sometimes give odd test results and unhealthy people can sometimes test normally. The important thing is that doctors use common sense - or additional clinical clues - to try to reason whether a person is truly damaged.

Second, these abnormal test results in apparently normal people are not entirely a random effect. People are more likely to score abnormally at least once if they take many tests. If the proper criteria for abnormality are imposed, the likelihood of abnormal tests scores also decreases.

That makes sense: Even if you take a test and get a hundred, that grade really isn't a truly accurate description of your knowledge because if you kept testing over and over, eventually you would probably get one wrong. The military follows this logic when testing applicants; if you get 50/50, your "real" score is 98/100. Hoorah!

Third, abnormal testing is not tied to gender in any statistically significant way. Men are not more likely to be "abnormal," nor is the reverse true, despite what you may want to think about your most recent ex. Race was also not a significant factor.

However, rates of abnormality do show interesting correlations to age, education and IQ levels.

But even after these variables are factored out of the equation, scientists have still found that certain levels of abnormal testing in healthy adults are "normal."

It is vital that doctors receive accurate tests for the normality of their patients' performances so they can make accurate and helpful diagnoses. Therefore, the team has concluded, there are certain statistical models that predict levels of abnormal test scores in healthy, normal adults.

But because the simulations are not always perfect, the best way of analyzing test results is this: If a patient tests normally more often than is predicted for normal people, that person is likely to be fine.

The opposite is also generally true: If a person tests significantly more abnormally than the predictions for a normal person, it might be that the person's brain is somehow not working at normal capacity.

But that's about as accurate and straightforward as the test results come. Mainly, it's important that doctors use their training and keen skills of observation.

The point to go home with is this: "Abnormality" rolls along a bell curve, so even with one little abnormality blighting your record, you're likely ahead of the curve, anyhow. This is Hopkins, after all.