Bayesianism helps but isn't taught well enough in school. Basically, we fail to handle the false positive and false negatives into the calculation, and this happens a lot with psychologists as well. This is really the point where people say untrained 'professionals' are dangerous - they can't evaluate that inaccuracy of the diagnosis itself.

This is the best explanation I've seen so far: https://betterexplained.com/articles/an-intuitive-and-short-...