Imagine you're a doctor in a town of 600 people facing an outbreak of a rare but deadly tropical disease. If the epidemic goes unchecked, the entire population will be wiped out.

You determine that there are only two possible ways of treating the epidemic.

If you order the first treatment, 200 people will be saved.

If you order the second, there's a one-third chance that all 600 will be saved and a two-thirds chance that none will be saved.

Which treatment should you choose?.

Three out of four people chose the first plan, in a study by two psychologists, Amos Tversky of Stanford University and Daniel Kahneman of the University of British Columbia.

Now consider another situation:

Again, you must choose between two ways of dealing with an epidemic.

But this time, if you order the first treatment, 400 people will die.

If you order the second, there's a one-third chance that no people will die and a two-thirds chance that all 600 will die.

Which plan should you choose?

This time, four out of five people chose the second approach.

The results are surprising because the probabilities in the two scenarios are identical. The only difference: One is worded in terms of lives saved and the other in terms of lives lost. "People are more willing to take a risk in order to avoid a sure loss," Tversky said. In the case above, hoping to avoid the sure loss of 400 lives by the first plan, they're willing to risk the uncertainty of the second plan.

But when the same choice is phrased in terms of a potential gain, as in the first scenario, people are less willing to take the risk, he said. They forgo the risk in favor of the certainty of saving 200 lives.

Similar confusion can arise when doctors inform patients about the relative risks of two different treatments, according to a study by Dr. Barbara McNeil of Harvard and her co-researchers, published in The New England Journal of Medicine.

People were asked to imagine they had lung cancer and had to choose between two therapies: surgery and radiation. The two therapies and their risks were described, including the odds of surviving one year and five years.

With surgery, one group was told, 10 percent of patients die during surgery and 66 percent die within five years. With radiation, no one dies during treatment but 78 percent die within five years. Another group was given the same odds expressed in terms of surviving instead of dying.

When the odds were presented in terms of surviving, 44 percent of the subjects chose radiation over surgery. But when the odds were presented in terms of dying, that percentage dropped to 18. The effect of the word change was equally strong among doctors and laypeople.