In order to use Bayesian reasoning to predict the probability that a patient sitting in a doctor's surgery has flu, we would need to know the proportion of people in the general population who show the same symptoms as the patient, the proportion of people in the population at large who have flu, the probability that the symptoms displayed by the patient do in fact indicate that the patient has flu, and the probability that the patient would show the given symptoms if it could be established that he did in fact have flu.

Which of the following piece of evidence would NOT be useful in establishing the necessary evidence that would allow Bayesian reasoning to take place regarding whether the patient had flu or not?


(A) A survey of 1000 randomly chosen people detailing how many of them had high temperatures and clogged noses, where these are known to be symptoms of flu.

(B) An investigation which revealed that 92% of the patients who have flu show the same set of symptoms as the patient currently under investigation.

(C) A study that indicated that at any given moment, approximately 95% of the general population do not suffer from flu.

(D) A medical text book that demonstrated that the symptoms mainly associated with flu could also be indicative of two other
diseases.

(E) The estimate by an expert in the field of contagious diseases that approximately 30% of the patients showing the same symptoms as the patient are in fact suffering from flu.
_________________

"Education is what remains when one has forgotten everything he learned in school."

A vs D.

I think it is D

I go with D.
_________________

SAID BUSINESS SCHOOL, OXFORD - MBA CLASS OF 2008

I think (A).The only one not taking care of the symptoms shown by the patient himself.
_________________

A well-balanced person is one who has a drink in each of his hands.

Straight D. All the others are related to atleast one part of the 'required' data.

D

D also