False Positive v. False Negative

by Dr. Lorin Bradbury

Question: Dr. Bradbury, can you explain the difference between a test that is prone to false positives vs. one that comes up with false negatives, and what is the advantage?

Answer: The first thing to be aware of is that any type of test produced by humans is not 100% accurate. All tests have potential for some degree of error. Let’s suppose someone could come up with a test that could determine whether a defendant was guilty without having to go through the cost of investigation and a trial. Knowing that all tests have some degree of error, you would want to develop a test that produced some false negatives, rather than false positives.

In other words, you would rather allow someone guilty to go free than you would to incarcerate or penalize someone who is actually not guilty. In that scenario, you would want to develop a test that erred on the side of producing false negatives.

In medical testing, such as for Covid-19, since you know all tests have potential for some error, you would want to develop a test that will identify all individuals who have Covid-19, but inadvertently identify a small percentage that are not infected with the virus. In this case, it would be better to develop a test that produced a small percent of false positive, rather than false negatives.

Often times screening tests are less intrusive and less sophisticated but produce higher rates of false positive results. They generally are quicker to administer and catch all positive cases, along with a few that are actually negative, but appear positive. That is why individuals who test positive with a screening test will be tested again with a test that is sometimes more intrusive, but more accurate.

Lorin L. Bradbury, Ph.D. is a licensed psychologist in private practice in Bethel. For appointments, he can be reached at 543-3266. If you have questions that you would like Dr. Bradbury to answer in the Delta Discovery, please send them to The Delta Discovery, P.O. Box 1028, Bethel, AK 99559, or e-mail them to [email protected].

Example: 9075434113