Here’s the picture of a plane with red dots in some area of its body. Basically, this is a bomber plane which had returned to the mission. The red dots are damaged point of the bomber plane itself. Every plane has different dots, and from the picture is the summary of damaged points.
The Researcher began to add the areas which have the most point or damaged. It seems logical because the researcher added armor to prevent the damaged area will damage in another time.
Until Abraham Wald, the statistician noted that the study of the plane which got damaged, only represent the plane which came back from a mission, not the plane which destroyed and unable to go back.
Wald then proposed that the Navy should take care of the areas which not damaged in the airplane. So if it hits to make the airplane down and unable to return.
Survivor bias based on the example above give us information about what should we learn about the selection bias. It makes people believe that you have to be the same factor as the survivor. It would ignore the failure one, so doesn’t want to learn from that factor.
If there are a people who do the certain way of education which make them a successful person, that doesn’t mean that their education is guaranteed to be a success. It needs to be checked with the people who failed even already took the same way of education. Why they failed and why the other is survived.
-18 August 2018-