When interacting with people who have lived to be 100, or even 110 years old or more, one often encounters the cliché and somewhat tedious questions like: “What did you do to live so long?”
Certainly, some interesting and surprising answers will emerge and capture attention. For example, the routine of eating fish and fries every Friday, or having a shot of hard liquor daily, or enjoying bacon at every breakfast, or continuously indulging in wine and chocolate.
While reporting on centenarians often attracts a wide readership, asking the questions mentioned above is relatively pointless and does not help us understand why some people live so long. In an article sent to CNN, author Brady Elliot from Westminster University in the UK argues why seeking the “secret” to longevity from long-lived individuals is futile.
The survivor bias can lead us to have an inaccurate view of how a person achieves longevity. (Source: CNN).
He points out that during World War II, statisticians from the Allied forces attempted to apply their skills to minimize the number of bombers shot down by enemy fire. By studying the damage patterns of bombers returning from missions, they mapped out which parts of the planes were most frequently damaged and recommended adding better, more expensive armor to these areas.
It sounds like a simple solution, right? However, that was when another statistician, Abraham Wald, intervened. He offered a completely opposite perspective, pointing out that the planes being studied had survived their missions despite sustaining heavy damage. But what about the planes that did not return?
Wald argued that armor should be added to the areas that showed no damage on all the planes that returned successfully. According to him, any plane that was hit in those undamaged areas might have crashed and failed to return to base for analysis.
Survivor Bias
The phenomenon that Wald countered is known as survivor bias, or cognitive and statistical bias created by only considering the factors that are present, while ignoring the factors that did not “survive.”
Try applying survivor bias to the study of longevity tips, and you will quickly see the absurdity. Imagine a group of 100 people, all of whom have smoked their entire lives.
As a group, regular smokers are likely to die earlier from cancer, lung disease, or heart disease. Yet within this group, there may still be one or two individuals who defy the trend and live to be 100 years old.
Now imagine journalists surrounding these lucky individuals on their 100th birthday, asking the classic question we all know: “What do you attribute your long life to?”
“I smoke a pack of cigarettes every day,” would likely be the response of the centenarian.
It may sound absurd, but survivor bias is prevalent throughout society. We have all heard stories of a famous actor or entrepreneur who has succeeded against the odds. They worked hard, believed in their personal goals, and succeeded. However, we rarely, if ever, hear the stories of those who tried, gave their all, but ultimately did not achieve success.
To be frank, writing or telling stories about failures is not great media material. But this contributes to the creation of survivor bias: we primarily hear success stories, while failures are seldom mentioned.
This bias applies and shapes our perceptions of architecture (most of us only know about the great buildings from a certain period), finance (we often hear examples of those who succeeded with risky investments, while failures don’t publish books or self-help courses), and career planning (you may have heard successful individuals recount how they dropped out of school and became wealthy, but you don’t hear about the many others who did the same and failed).
Brady Elliot shares that he has worked with many older adults, including extremely elderly individuals. He and his colleagues at the university are studying those over 65 who maintain an unusually high level of physical activity in old age and enjoy excellent health.
They are exceptional examples of older adults who are still faster and healthier than Elliot, despite being nearly double his age. However, even though they know that lifelong exercise is associated with the good health of these individuals in old age, Elliot’s team still cannot draw specific conclusions about which led to which. It could be that regular exercise protects centenarians from serious diseases like cancer, diabetes, and heart disease. But it could also be that these individuals continue to exercise and maintain good health in old age because they never had cancer, diabetes, or heart disease earlier in life.
There may also be some unidentified third factors that scientists still have to discover to understand the mechanisms that allow many people not only to live long but also to maintain their ability to exercise for good health.
Correlation does not necessarily imply causation. This point is often emphasized to students conducting scientific research. Our brains often see similar patterns between two variables and assume they are related in some way. But often, like survivor bias, we do not consider enough input data, so we “detect” a correlation where there may actually be none.
