Probability quiz

[OohPooPahDoo]

POLL QUIZ:

Suppose a test to determine if a patient has a certain disease has a 5% false positive rate. (0% false negative). Suppose the chances of someone selected at random of having the disease is 0.1%. You select someone at random from the population, administer the test, and it comes back positive. What are the chances this person actually has the disease?

 

[rightwinger]

Where do you bury the survivors?

 

[Rozman]

POLL QUIZ:

Suppose a test to determine if a patient has a certain disease has a 5% false positive rate. (0% false negative). Suppose the chances of someone selected at random of having the disease is 0.1%. You select someone at random from the population, administer the test, and it comes back positive. What are the chances this person actually has the disease?

Welcome to Florida! Someone look funny to you? Stalk them and shoot them to death, its 100% legal.
Reply With Quote

When did this happen?.....

 

[Rozman]

Suppose a test to determine if a patient has a certain disease has a 5% false positive rate. (0% false negative). Suppose the chances of someone selected at random of having the disease is 0.1%. You select someone at random from the population, administer the test, and it comes back positive. What are the chances this person actually has the disease?

Is this a brain teaser...



A train traveling from Boston to NY stops at the first station and 25 people get on and 13 get off.
the next stop 53 get on and 23 get off...The next stop 47 get on and 22 get off....The train finally stops at NY....

Who won the Yankee Red Sox game....?

 

[BobPlumb]

I have the first purple answer.

 

[OohPooPahDoo]

I have the first purple answer.

Well done BobPlumb.


BobPlumb has the correct answer.

There are 1000 people. The test has a 5% false positive rate. That means out of 1000 people, 50 will test positive but not have the disease. But only 0.1% - or 1 in 1000 - actually have the disease. So for every person with the disease, there are 50 people who will test false positive. Thus the total odds of a randomly selected person who tests positive actually having the disease is 1 in 51 - or just under 2%.


Unfortunately most doctors probably would have answered 95%.

This illustrates the problem with doctors administering tests blindly - unless the disease they are testing for is very common and/or the test is extremely accurate, there's a good chance a positive result is relatively meaningless. In the above example, a positive result means you go from a 0.1% chance to a 1.91% chance of having the disease - not entirely insignificant - but not enough to base a program of treatment on.

 

[CrusaderFrank]

...and this is in "Stock Market"??????

 

[OohPooPahDoo]

...and this is in "Stock Market"??????

The stock market has everything to do with probability.

This quiz shows that most people don't understand it.


For instance - what do you think of the statement "Most options expire worthless - thus it is better to be an option seller than option buyer"?

 

[CrusaderFrank]

...and this is in "Stock Market"??????

The stock market has everything to do with probability.

This quiz shows that most people don't understand it.

For instance - what do you think of the statement "Most options expire worthless - thus it is better to be an option seller than option buyer"?

The odds are good you're going to go off on a tangent

 

[OohPooPahDoo]

...and this is in "Stock Market"??????

The stock market has everything to do with probability.

This quiz shows that most people don't understand it.

For instance - what do you think of the statement "Most options expire worthless - thus it is better to be an option seller than option buyer"?

The odds are good you're going to go off on a tangent

lol


The statement in and of itself is worthless. 9 in 10 options may expire worthless but what we need to know to answer the question is what will the 10th option pay out?



Here's another one - "I believe the market is most likely to go down tomorrow. However, its in my best interest to buy stock today"

Can that statement be logically consistent? Yes. If chances are 2 in 3 the market will go down by 1% - but there is a 1 in 3 chance it will go up by 5% - you get higher expected returns by buying stock.

Most players in the market have trouble understanding skewed probability distribution. Most scientists and doctors and people in general do.


EDIT:
I have pulled some of these ideas from "Fooled by Randomness" by Nassim Taleb. Its a great read.

 

[CrusaderFrank]

The stock market has everything to do with probability.

This quiz shows that most people don't understand it.

For instance - what do you think of the statement "Most options expire worthless - thus it is better to be an option seller than option buyer"?

The odds are good you're going to go off on a tangent

lol


The statement in and of itself is worthless. 9 in 10 options may expire worthless but what we need to know to answer the question is what will the 10th option pay out?


Here's another one - "I believe the market is most likely to go down tomorrow. However, its in my best interest to buy stock today"

Can that statement be logically consistent?

I like the Graham/Buffett Concept of "Mr. Market", some days he's very happy and optimistic, other days he's depressed and grumpy

 

[BobPlumb]

I have the first purple answer.

Well done BobPlumb.


BobPlumb has the correct answer.

There are 1000 people. The test has a 5% false positive rate. That means out of 1000 people, 50 will test positive but not have the disease. But only 0.1% - or 1 in 1000 - actually have the disease. So for every person with the disease, there are 50 people who will test false positive. Thus the total odds of a randomly selected person who tests positive actually having the disease is 1 in 51 - or just under 2%.


Unfortunately most doctors probably would have answered 95%.

This illustrates the problem with doctors administering tests blindly - unless the disease they are testing for is very common and/or the test is extremely accurate, there's a good chance a positive result is relatively meaningless. In the above example, a positive result means you go from a 0.1% chance to a 1.91% chance of having the disease - not entirely insignificant - but not enough to base a program of treatment on.

Do I get a cookie?

 

[Leweman]

Mind blowing stuff. How do you ever know that .1% of people have it though?

 

[konradv]

...and this is in "Stock Market"??????

The stock market has everything to do with probability.

This quiz shows that most people don't understand it.

For instance - what do you think of the statement "Most options expire worthless - thus it is better to be an option seller than option buyer"?

The odds are good you're going to go off on a tangent

Yeah, don't you know that's Frank's thing?!?!

 

[rightwinger]

POLL QUIZ:

Suppose a test to determine if a patient has a certain disease has a 5% false positive rate. (0% false negative). Suppose the chances of someone selected at random of having the disease is 0.1%. You select someone at random from the population, administer the test, and it comes back positive. What are the chances this person actually has the disease?

Regardless of the test results, the chances of someone having the disease remains at 0.1%

The test results do not change your probability of having the disease only the probability of detecting it correctly.

 

[konradv]

POLL QUIZ:

Suppose a test to determine if a patient has a certain disease has a 5% false positive rate. (0% false negative). Suppose the chances of someone selected at random of having the disease is 0.1%. You select someone at random from the population, administer the test, and it comes back positive. What are the chances this person actually has the disease?

Regardless of the test results, the chances of someone having the disease remains at 0.1%

The test results do not change your probability of having the disease only the probability of detecting it correctly.

We're not talking about "someone". We're about a particular person with a test that comes back positive. If the test is 95% accurate, then that's the chance they actually have the disease. It's not 0.1% because the case is no longer random. We have an actual test that tells us the probability.

 

[Leweman]

POLL QUIZ:

Suppose a test to determine if a patient has a certain disease has a 5% false positive rate. (0% false negative). Suppose the chances of someone selected at random of having the disease is 0.1%. You select someone at random from the population, administer the test, and it comes back positive. What are the chances this person actually has the disease?

Regardless of the test results, the chances of someone having the disease remains at 0.1%

The test results do not change your probability of having the disease only the probability of detecting it correctly.

We're not talking about "someone". We're about a particular person with a test that comes back positive. If the test is 95% accurate, then that's the chance they actually have the disease. It's not 0.1% because the case is no longer random. We have an actual test that tells us the probability.

Not really because there are so many people out of a thousand that don't have it. So if you go with the 1 out of a thousand having it that means 50 out of a thousand will get a positive but don't have it. Then you add the 1 that will test positive and actually has it. Thats 51 people that will test positive but only one person will actually have it ( 1 out of 51) That is 2% roughly. I still don't understand how you ever get to the point of knowing what percentage of the population actually has it though if the tests aren't accurate.

 

[BobPlumb]

When a hypothetical math problem is made up, the writer of the problem is given the information from the math gods.

 

[jwoodie]

regardless of the test results, the chances of someone having the disease remains at 0.1%

the test results do not change your probability of having the disease only the probability of detecting it correctly.

we're not talking about "someone". We're about a particular person with a test that comes back positive. If the test is 95% accurate, then that's the chance they actually have the disease. It's not 0.1% because the case is no longer random. We have an actual test that tells us the probability.

not really because there are so many people out of a thousand that don't have it. So if you go with the 1 out of a thousand having it that means 50 out of a thousand will get a positive but don't have it. Then you add the 1 that will test positive and actually has it. Thats 51 people that will test positive but only one person will actually have it ( 1 out of 51) that is 2% roughly. I still don't understand how you ever get to the point of knowing what percentage of the population actually has it though if the tests aren't accurate.

f