The choice was between two treatments for lung cancer, and the numbers were laid out plainly.
Of 100 people having surgery, 90 survive the treatment, 68 are still alive after one year, and 34 after five years. Of 100 people having radiation therapy, all survive the treatment, 77 are alive after one year, and 22 after five years.
Surgery carries a risk up front and leaves more people alive at five years. Radiation is safer at the start and leaves fewer. Reasonable people can choose either.
Other people read the same choice told the other way. Of 100 people having surgery, 10 die during treatment, 32 will have died by one year, and 66 by five years. Of 100 having radiation, none die during treatment, 23 will have died by one year, and 78 by five years.
Nothing in the second version is new. Ninety alive is ten dead.
The study, published in the New England Journal of Medicine in May 1982, came from three physicians, Barbara McNeil, Stephen Pauker and Harold Sox, and the psychologist Amos Tversky. They gave versions of the choice to 238 outpatients with various chronic illnesses, 491 graduate students and 424 physicians. Everyone was asked to imagine having lung cancer. Some versions named the treatments and some didn’t; some gave survival over time and some gave life expectancy; and some were told in terms of living, others in terms of dying.
In all three groups, surgery looked substantially more attractive when the problem was framed in terms of living than in terms of dying. The physicians were no exception.
As Laura Macchi and Edoardo Zulato summarized the original results in 2021, 75 percent chose surgery when the outcomes were about surviving, and 58 percent when they were about dying. By McNeil and colleagues’ interpretation, as Macchi and Zulato describe it, the 10 percent risk of dying in surgery set against radiation’s zero loomed larger than 90 percent surviving set against 100.
Macchi and Zulato reran the original vignettes with 150 students of medicine, statistics and other subjects. The gap came out even wider: 81.3 percent chose surgery in the survival version, 20.0 percent in the mortality version.
Then they changed one more thing. In the original, the numbers pile up: the 32 who have died by one year include the 10 who died during treatment. Macchi and Zulato suspected that readers of the dying version added the numbers up as if each were new deaths. When the same outcomes were written as separate counts for each period (10 die during treatment, 22 more by one year, 34 more by five years), the framing effect in their study largely disappeared.