Chris Newton / Getty Images stock

“A physician who treats himself has a fool for a patient,” Sir William Osler, a founding faculty member of Johns Hopkins Hospital and its famous school of medicine, once said. Yet we often try to diagnose ourselves using Dr. Google.  

Now, new research from the Hong Kong University of Science and Technology helps explain why that’s a very bad idea: The closer we are to the subject -- and when we’re diagnosing ourselves we’re very close indeed -- the more likely we are to over- or under-estimate the chances we’ve got something.

Marketing PhD candidate Dengfeng Yan, who will assume a teaching post at the University of Texas at San Antonio this fall, and his department chairman, Jaideep Sengupta, conducted a series of experiments showing how our cognitive brains are subject to biases that affect judgment of disease risk.  

Building on past research, they explored self-positivity and self-negativity biases. Such research has shown, for example, that symptoms we perceive to be indigestion in a stranger are often thought to be a possible heart attack in ourselves (self-negativity). Conversely, we can underestimate our risk for many common conditions like sexually transmitted diseases (self-positivity).

Using hundreds of university students, they presented a variety of scenarios involving diseases like flu, hepatitis C, breast cancer, osteoporosis. They provided different sets of information (high or low) on the “base-rate” -- meaning the incidence of a condition in the population -- and “case-risk” -- one person’s profile of behaviors or symptoms. Sometimes the person was a stranger, sometimes themselves.

The experiments showed that social distance mattered. The less familiar the person in the scenario was, the more heavily the test subjects relied on base-rate information. The closer to the subjects, including themselves, the more they relied on individual case information.

“We found the effect to be quite strong, as evidenced by the fact that we replicated our findings using different manipulations of psychological distance, and across five different types of health risks,” Yan told NBC News.

For example, subjects were given differing sets of data about the rate of HIV in Hong Kong, and then given case information including a scenario about themselves or about a stranger. They were asked “How likely are you [or how likely is the stranger] to engage in risky behaviors by which HIV is transmitted?”

When told the disease base rate was high, but risky behavior likelihood was low, the subjects said they were less vulnerable to risk than others (who engaged in the same behaviors) were. That’s a self-positivity bias. But when told the base rate of HIV was low, but risky behavior likelihood was high, the participants judged themselves more vulnerable than others. That’s self-negativity.

Another test used two strains of flu, one mild and one more dangerous. Again, they tended to either underestimate their own risk when told that the base rate of mild flu was high but they had only one symptom, and overestimate their risk for dangerous swine flu when told the base rate was low, but they themselves had multiple symptoms, though those symptoms could also apply to the common cold or allergies.

Though he tested people of undergraduate age, Yan believes his results would hold for people of any age.

His findings, published this month in the Journal of Consumer Research, show that the advantage of seeing a real doctor isn’t just because he or she is an expert. It’s also that they aren’t you.      

Brian Alexander (www.BrianRAlexander.com) is co-author, with Larry Young PhD., of "The Chemistry Between Us: Love, Sex and the Science of Attraction," (www.TheChemistryBetweenUs.com) to be published Sept. 13.

Related:

• When you can't get over your bridge phobia
• Why you can't get 'Call Me Maybe' out of your head
• Phew! The science of the close call