I am going to have to give the answer before Ed pulls out the last of his rapidly thinning locks and finally gets sick of PM'ing me for the answer.
The correct answer is 2%
You have been diagnosed with a disease that has a prevalence of 1 in a 1000 in the community. The test has an accuracy of 95%. Thus 50 people in every 1000 (5%)will be diagnosed with the disease but 49 will be false positives leaving the chances of you having the disease 1 in 50 or 2%
A man was out camping. He left his tent in the morning to take a dump and first walked 500 metres south. After finishing up he walked 500 metres east, turned and walked another 500 metres north.
He then saw a bear ransacking his tent. What colour was the bear?
^
That'll be white Mendy, I remember that one from playschool
......
50 people get a false result how can you predict whether a false positive or false negative without more data, I think your parameters contain a category error[/QUOTE]
Sorry David no error,
For every 1000 people tested statistically 50 will be diagnosed with the disease, as the test is only 95% accurate and the prevalence is 1 in 1000 the chances of you having the the disease is 1 in 50
It's basic statistics.
but you have not refuted this.
You may be right but how can you know
1 if the false results were false - or +?
Your data gives no clue
2 the disatribution between false positive and false positive?
In your model can you tell me how many were
False Positive
and how many were False Negatives?
I think not
^ David statistically speaking it dictates that 49 out of 1000 tests will be false positives that leaves 1 out of the 1000 who should have the disease that is 1 from 50 ie 2%
So we are left with 49 false positives and 1 positive.
Are you taking the piss?
Not taking the piss and you may be right , but you still dont explain how many false negatives do you, or have I missed something. Anyway put up another I need to oil my old brain, perhaps you guys are too smart for me or I misread the OP
So how many FALSE POSTIVES is my question
^ Read my post it clearly states 49 false positives from 1000.
Let's get over this one soon, once you all catch up I will post a real contentious head scratcher that will cause great arguments TD style
A particular heart disease has a prevalence of 1/1000 people
Does this link help?
Sorry David no error,
For every 1000 people tested statistically 50 will be diagnosed with the disease, as the test is only 95% accurate and the prevalence is 1 in 1000 the chances of you having the the disease is 1 in 50
It's basic statistics.[/QUOTE]
You got your basics statistics wrong.
Google might be your friend on this problem. Look it up and check it out.
The extent of the testing makes all the difference. Google is not always your friend
There are currently 1 users browsing this thread. (0 members and 1 guests)