Comparing percents?

• May 10th 2008, 10:11 AM
TK_M
Comparing percents?
I keep seeing that one in five of us will die from cancer.

So in a family of five, each of us has a 20% chance of dying from it individually, but what is the chance of the family as a whole having someone dying from it? Clearly it is not 100% as another family of five might have two die in it and you have none.

I am ignoring the heredity factor in some types of cancer, just assume the cases happen at random.

I would be grateful for any help as this sort of problem has been puzzling me for a while!
• May 10th 2008, 10:54 AM
janvdl
Quote:

Originally Posted by TK_M
I keep seeing that one in five of us will die from cancer.

So in a family of five, each of us has a 20% chance of dying from it individually, but what is the chance of the family as a whole having someone dying from it? Clearly it is not 100% as another family of five might have two die in it and you have none.

I am ignoring the heredity factor in some types of cancer, just assume the cases happen at random.

I would be grateful for any help as this sort of problem has been puzzling me for a while!

You have to realise that the family's habits will obviously influence this.

Staying out of the sun will obviously decrease your chances of skin cancer, and eating correctly and taking in all the necessary minerals and vitamins, will also decrease your chances of getting another cancer in turn.

However the chances of 1/5 that someone in your family has cancer is not very far-fetched. In my own family we have had our fair share of cancer patients. ( Although it would rather be about 10% than 20% )
• May 10th 2008, 11:00 AM
janvdl
Quote:

Originally Posted by TK_M
But what is the chance of the family as a whole having someone dying from it? Clearly it is not 100% as another family of five might have two die in it and you have none.

I'm not sure what you mean by that exactly. The average family will have 1 in 5 of its members die from cancer.

The chances of your family having no deaths from cancer will probably follow a Poisson distribution.

Take the average as 1/5
Chances of 0 deaths if following a Poisson distribution: 81,87%
• May 12th 2008, 05:04 AM
TK_M
Quote:

Originally Posted by janvdl
I'm not sure what you mean by that exactly. The average family will have 1 in 5 of its members die from cancer.

The chances of your family having no deaths from cancer will probably follow a Poisson distribution.

Take the average as 1/5
Chances of 0 deaths if following a Poisson distribution: 81,87%

Thank you Janvdl! This sounds like the sort of value I would have expected.

Did you mean the chance of one death in a family of 5 is 81.87%, not 0 deaths is 81.87% though? Only 18.13% seems a bit low?

Sorry for not being clearer with my question. I was asking if 1 in 5 people die from cancer, what is the chance of 1 or more deaths in a family of 5? (assuming all cases of cancer happen randomly, that is not inherited, or due to environmental factors).

Generally, you could state my problem as:
I have an unknown quantity of marbles in a bag. 1 in 5 is red, the rest are blue. If I pick 5 marlbes out of the bag randomly, what is the chance that at least one of them will be red?

Thank you so much for your help!

Tony
• May 12th 2008, 05:31 AM
janvdl
Quote:

Originally Posted by TK_M
Thank you Janvdl! This sounds like the sort of value I would have expected.

Did you mean the chance of one death in a family of 5 is 81.87%, not 0 deaths is 81.87% though? Only 18.13% seems a bit low?

Sorry for not being clearer with my question. I was asking if 1 in 5 people die from cancer, what is the chance of 1 or more deaths in a family of 5? (assuming all cases of cancer happen randomly, that is not inherited, or due to environmental factors).

Generally, you could state my problem as:
I have an unknown quantity of marbles in a bag. 1 in 5 is red, the rest are blue. If I pick 5 marlbes out of the bag randomly, what is the chance that at least one of them will be red?

Thank you so much for your help!

Tony

The chances that no-one in your family of 5 will die of cancer is 81,87%

Okay so if the average is 1/5, we could work out a Poisson probability for 1 death.
Which is: 16,37%