Why silence in data is not proof of absence
Launch library · evergreen read
When a dataset shows no recorded cases of something, that can mean the thing genuinely does not happen, or it can mean nobody was measuring it, nobody reported it, or the method used was not capable of detecting it at all, and these possibilities are easy to confuse with each other.
Absence of evidence becomes meaningful only once you know how carefully and how completely the relevant thing was actually looked for in the first place. A gap in the data from a system that never collected the relevant information tells you almost nothing about whether the underlying issue exists.
Careful communicators distinguish between an issue that has genuinely been studied and found absent, and one that has simply never been properly measured at all, since treating the second as though it were the first can quietly mislead an audience without anyone actually intending it, simply through an honest failure to look closely enough.