Much of what he »hear about statistics… is malformed. But Doctors, who we expect to understand the nature of tests, often do not.
Additionally, at the order of a single individual? Probability does not work. Probability requires a cohort, and the character of the cohort disappears with a sample size of 1. This is why doctors who employ statistics in client explanations are often completely wrong. No one can predict ‘at the level of a single individual’ with any dependable degree of probabilistic accuracy. While there may be a few explicit exceptions, in general, probability is about cohorts, and no person comprises a cohort (group). Testing, however, is a unique context, since we are testing for the presence of evidence of disease. This doesn’t tell us the outcomes, but is reasonably dependable (statistically) when the statistics are properly understood.
0 Comments