Introduction to Bayesian Thinking: from Bayes theorem to Bayes networks

Felipe Sanchez

6985

Here’s my problem: you stipulate that the test is 99% accurate and then you tell me that the chance that the test is, in fact, accurate in this case is only 9%. You give me a formula which you expect me to accept is valid but without giving me any reason to believe that it is.

You don’t explain why, if a 99% accurate test yields Result A, the odds that Result A is valid aren’t similarly 99%. You want me to accept your formula on faith.

You need to explain logically and convincingly why a 99% accurate test is actually 91% inaccurate, why the accuracy of a 99% accurate test is only 9% in any given case.