First-Hand Experience is Less Biased Than News

Source: EconLog
by Bryan Caplan

“One of the first lessons you learn in statistics is to discount first-hand experience. ‘But I knew a guy who …’ is weak evidence, for a long list of reasons …. While these are all reasonable concerns, they dodge a critical question: ‘Compared to what?’ When someone sets aside their first-hand experience, statisticians are hoping — perhaps even assuming — that they rely on random sampling instead. In the real world, however, almost no one does this. For the vast majority of human beings, the alternative to first-hand experience is not statistics, but news. And compared to news, first-hand experience is ultra-reliable, for a long list of reasons.” (07/09/21)