Is a geometric distribution a reasonable probability model for this case?
Suppose that one of every 100 people in a certain community is infected with HIV. You want to identify an HIV-positive person to include in a study of an experimental new drug.
Explain why or why not a geometric distribution is reasonable probabilty model for this case? Give your reasons
Also, if you can...How many individuals would you expect to have to interview in order to find the first peron who is HIV positive?
2007-01-01 12:24:24 · 2 answers · asked by Anonymous in Science & Mathematics ➔ Mathematics
Yes the geometric distribution is a fair assumption. With there being only an underlying probability of success (finding someone with HIV) the other most logical test (at least in my mind) would be binomial. However it is possible to have 0 people within the sample of a binomial trial whereas a geometric has no set number of trials.
Since P(HIV) = 1/100 = p
E(X | X~ Geometric) = 1/p = 1/(1/100) = 100
2007-01-01 12:39:27 · answer #1 · answered by Modus Operandi 6 · 0⤊ 0⤋
Go get permission from the FDA. They will give you a list, with the patient's consent of who has HIV; In confidence of course. I think.
2007-01-01 20:29:52 · answer #2 · answered by Anonymous · 0⤊ 1⤋