Ask your own question, for FREE!
Mathematics 8 Online
OpenStudy (anonymous):

Question… In a group of 100 people, 20 are albino. If I choose 5 people at random, what is the chance: 0 are albino 1 is albino 2 are albino 3 are albino 4 are albino 5 are albino

OpenStudy (anonymous):

Am I correct to say, that all the above cases should add up to a probability of 1? What I have: for 0 albino: 80/100*79/99*78/98*77/97*76/96 = 0.3193… for 1 albino: 20/100*80/99*79/98*78/97*77/96 = 0.0840… for 2 albino: 20/100*19/99*80/98*79/97*78/96 = 0.0207… you get the picture. When I add them up, I don't get anything near 1 (or even 0.5)

OpenStudy (anonymous):

1 in 5 are albino 5 are randomly chosen 0 are albino, 4/5 1 albino 1/5 2,3,4,5 albino = no chance

OpenStudy (anonymous):

i think

OpenStudy (kropot72):

\[P(0)=\frac {\left(\begin{matrix}20 \\ 0\end{matrix}\right)\left(\begin{matrix}80 \\ 5\end{matrix}\right)}{\left(\begin{matrix}100 \\ 5\end{matrix}\right)}\]

OpenStudy (kropot72):

\[P(1)=\frac{\left(\begin{matrix}20 \\ 1\end{matrix}\right)\left(\begin{matrix}80 \\ 4\end{matrix}\right)}{\left(\begin{matrix}100 \\ 5\end{matrix}\right)}\]

OpenStudy (anonymous):

Do you know what was wrong with my method? I can't see it

OpenStudy (kropot72):

This problem is sampling without replacement. The hypergeometric disgtribution applies. \[P(2)=\frac{\left(\begin{matrix}20 \\ 2\end{matrix}\right)\left(\begin{matrix}80 \\ 3\end{matrix}\right)}{\left(\begin{matrix}100 \\ 5\end{matrix}\right)}\]

OpenStudy (anonymous):

sorry, i don't understand

OpenStudy (kropot72):

*distribution

OpenStudy (amistre64):

kropot took my idea, the hypergeometric distribution

OpenStudy (anonymous):

can you please explain the hypergeometric distribution

OpenStudy (kropot72):

@psujono Do you understand the binomial coefficient\[\left(\begin{matrix}n \\ r\end{matrix}\right)=\frac{n!}{r!(n-r)!}\]

OpenStudy (anonymous):

ya

OpenStudy (kropot72):

Well you can calculate the values of P(0), P(1) and P(2) from the equations I have already given.

OpenStudy (anonymous):

i love the word "hyper-geometric"

OpenStudy (anonymous):

Oh yeah, sure thanks... I did a little experiment, it turns out you get the same results if you had a group of 50, and 10 albino. (obviously) Could we generalise this into a formula such: You have that x/y (<-- fraction in simplest form) are albino And we want to pick a group of n, where p are albino...

OpenStudy (kropot72):

Sorry to say I think it is a horrible word. Just about as bad as the word statistics :(

OpenStudy (anonymous):

@psujono there is nothing wrong with your method, so long as you are careful

OpenStudy (anonymous):

^ actually, not so obviously it turns out I was wrong. The percentages change slightly. @satellite73, how come in my first post I got such different answers?

OpenStudy (anonymous):

take for example your calculation of the probability you get two albino you write 20/100*19/99*80/98*79/97*78/96 which makes sense, but this is the probability you pick the two albino first

OpenStudy (anonymous):

to adjust, multiply by the numbers of ways to arrange the two albino in the 5 picks, namely \(\binom{5}{2}=10\)

OpenStudy (anonymous):

For 2 albino, I got 0.0207... with @kropot72's method, I got 0.2073...

OpenStudy (anonymous):

How come this is necessary?

OpenStudy (anonymous):

right, as i said, you computed "albino, albino, not, not, not" in that order

OpenStudy (anonymous):

Oh.. I get it :D

OpenStudy (kropot72):

P(0) = 0.3193 P(1) = 0.42 P(2) = 0.207 P(3) = 0.0478 P(4) = 0.00515 P(5) = 0.0002

OpenStudy (anonymous):

but there are ten different ways to arrange those two albinos all with the same probability so your method is intuitively correct, you just need to multiply by the possible number of ways to arrange each outcome

OpenStudy (anonymous):

Sure thanks!

OpenStudy (anonymous):

compute for example \[\frac{\dbinom{20}{3}\times \dbinom{80}{3}}{\dbinom{100}{5}}\] and \[\dbinom{5}{2}\times \frac{20}{100}\times \frac{19}{99}\times \frac{80}{98}\times \frac{79}{97}\times \frac{78}{96}\] without a calculator. i.e. by factoring and canceling you will see not only that you get the same thing, but maybe you will get some insight in to why you get the same thing (maybe)

OpenStudy (anonymous):

ok that way a typo, first line should have been \[\frac{\dbinom{20}{2}\times \dbinom{80}{3}}{\dbinom{100}{5}}\]

OpenStudy (anonymous):

Great, I see why it works now. Thanks

OpenStudy (anonymous):

yw

OpenStudy (anonymous):

so when 1 in 4 cereal boxes have a token, and there are 10000 cereal boxes, If you buy 4 boxes, there is a 68.37% chance you will get at least 1 token. How Insightful!

OpenStudy (anonymous):

really?

Can't find your answer? Make a FREE account and ask your own questions, OR help others and earn volunteer hours!

Join our real-time social learning platform and learn together with your friends!
Can't find your answer? Make a FREE account and ask your own questions, OR help others and earn volunteer hours!

Join our real-time social learning platform and learn together with your friends!