Survivorship bias is a tricky fallacy to deal with.
We’re all mesmerised by success stories. But studying those who are successful isn’t always the best idea.
You might succumb to survivorship bias.
You might bet on flawed strategies.
Here goes:
Abraham Wald and the Allied Aircrafts
During World War II, the Allies studied Nazi damage to their fighter planes. Their study resulted in this dotted illustration:
The analysis of these damages resulted in the idea that the Allied Forces should reinforce their fighter planes in the areas showing the densest clusters of red dots.
A statistician by the name of Abraham Wald disagreed. Instead, he suggested that they reinforce surfaces with no red dots. Wald was right, of course. The red-dotted areas indicated non-fatal damaged regions by studying planes that somehow returned home.
If it had been possible to study all the planes shot down and destroyed by the Nazis, maybe the research team would have found an inverse pattern?
“Wald’s work on aircraft vulnerability, using data from survivors, was crucial in World War II, Korea, and Vietnam, and has been reissued by the Center for Naval Analyses.”
Source: Journal of the American Statistical Association 1Mangel, M., & Samaniego, F. (1984). Abraham Wald’s Work on Aircraft Survivability. Journal of the American Statistical Association, 79, 259 – 267. https://doi.org/10.1080/01621459.1984.10478038
World War I and the Soldiers’ Helmets
In another example, we focus on a seemingly counterintuitive phenomenon that emerged during World War I.
As military forces began to increase the issuing of helmets to soldiers to protect them from the perils of the battlefield, an unexpected trend was observed: the number of wounded soldiers surged dramatically.
At first glance, it may seem that the helmets were paradoxically making it easier for more people to get hurt. The crux of this paradox lies in the shifting dynamics of the casualty statistics:
As helmets were introduced and distributed more widely, they significantly reduced the number of fatalities on the battlefield. Soldiers who would have previously succumbed to their injuries were now surviving. Consequently, these soldiers were classified as “wounded” rather than “dead.”
This shift in categorisation created the illusion of increased injuries when, in reality, the number of fatalities had decreased.
The Fallacy of Survivorship Bias
These examples underscore the importance of considering the larger context and being cautious about conclusions based solely on superficial data.
Both stories are excellent examples of a particular fallacy — survivorship bias.
“Survivorship bias, survival bias is the logical error of concentrating on entities that passed a selection process while overlooking those that did not. This can lead to incorrect conclusions because of incomplete data.”
Source: Wikipedia 2Survivorship bias. (2023, October 9). In Wikipedia. https://en.wikipedia.org/wiki/Survivorship_bias
We often can’t help being enticed by success stories in popular culture and business. But their stories can lead us to fallacious conclusions.
Few think to study those who turned out to be unsuccessful. This is too bad for data sets since the unsuccessful typically outnumber the successful by a vast margin.
Please support my blog by sharing it with other PR- and communication professionals. For questions or PR support, contact me via jerry@spinfactory.com.
PR Resource: Logical Fallacies and Biases
List of Logical Fallacies and Biases
As humans, we often fall for the tricks our own psychology plays on us. These “thinking errors” exist because they’ve often aided our survival. However, knowing and understanding various types of common fallacies and biases is helpful in everyday life.
Here are a few examples of logical fallacies and biases that I’ve come across while studying public relations and linguistics:
Learn more: 58 Logical Fallacies and Biases
💡 Subscribe and get a free ebook on how to get better PR ideas.
ANNOTATIONS
1 | Mangel, M., & Samaniego, F. (1984). Abraham Wald’s Work on Aircraft Survivability. Journal of the American Statistical Association, 79, 259 – 267. https://doi.org/10.1080/01621459.1984.10478038 |
---|---|
2 | Survivorship bias. (2023, October 9). In Wikipedia. https://en.wikipedia.org/wiki/Survivorship_bias |