In: Statistics and Probability
You roll two fair dice, and denote the number they show by X and Y. Let U = min{X, Y } and V = max{X, Y }. Write down the joint probability mass function of (U, V ) and compute ρ(U, V ) i.e the correlation coefficient of U and V
Following is the sample space of the rolling of two fair dice:
U shows the minimum of two dice. For example: for outcome (1,2) , U is 1.
V shows the maximum of two dice. For example: for outcome (1,2) , V is 2.
Since there are two outcome (1,2) and (2,1) for which minimum is 1 and maximum is 2. So
P(U=1, V=2) = 2/36
Likewise following table shows the joint pdf of U and V:
U | ||||||||
1 | 2 | 3 | 4 | 5 | 6 | P(V=v) | ||
1 | 1/36 | 0 | 0 | 0 | 0 | 0 | 1/36 | |
2 | 2/36 | 1/36 | 0 | 0 | 0 | 0 | 3/36 | |
V | 3 | 2/36 | 2/36 | 1/36 | 0 | 0 | 0 | 5/36 |
4 | 2/36 | 2/36 | 2/36 | 1/36 | 0 | 0 | 7/36 | |
5 | 2/36 | 2/36 | 2/36 | 2/36 | 1/36 | 0 | 9/36 | |
6 | 2/36 | 2/36 | 2/36 | 2/36 | 2/36 | 1/36 | 11/36 | |
P(U=u) | 11/36 | 9/36 | 7/36 | 5/36 | 3/36 | 1/36 | 1 |
---------------------------
Now we need to find the mean and variance of U and V and E(UV). Following table shows the calculations for E(UV):
U | V | P(U=u, V=v) | U*V*P(U=u, V=v) |
1 | 1 | 1/36 | 1/36 |
1 | 2 | 2/36 | 4/36 |
1 | 3 | 2/36 | 6/36 |
1 | 4 | 2/36 | 8/36 |
1 | 5 | 2/36 | 10/36 |
1 | 6 | 2/36 | 12/36 |
2 | 1 | 0 | 0 |
2 | 2 | 1/36 | 4/36 |
2 | 3 | 2/36 | 12/36 |
2 | 4 | 2/36 | 16/36 |
2 | 5 | 2/36 | 20/36 |
2 | 6 | 2/36 | 24/36 |
3 | 1 | 0 | 0 |
3 | 2 | 0 | 0 |
3 | 3 | 1/36 | 9/36 |
3 | 4 | 2/36 | 24/36 |
3 | 5 | 2/36 | 30/36 |
3 | 6 | 2/36 | 36/36 |
4 | 1 | 0 | 0 |
4 | 2 | 0 | 0 |
4 | 3 | 0 | 0 |
4 | 4 | 1/36 | 16/36 |
4 | 5 | 2/36 | 40/36 |
4 | 6 | 2/36 | 48/36 |
5 | 1 | 0 | 0 |
5 | 2 | 0 | 0 |
5 | 3 | 0 | 0 |
5 | 4 | 0 | 0 |
5 | 5 | 1/36 | 25/36 |
5 | 6 | 2/36 | 60/36 |
6 | 1 | 0 | 0 |
6 | 2 | 0 | 0 |
6 | 3 | 0 | 0 |
6 | 4 | 0 | 0 |
6 | 5 | 0 | 0 |
6 | 6 | 1/36 | 36/36 |
Total | 441/36 |
So,
---------------------
Following table shows the calculations for mean and var of U:
U | P(U=u) | u*P(U=u) | u^2*P(U=u) |
1 | 1/36 | 1/36 | 1/36 |
2 | 3/36 | 6/36 | 12/36 |
3 | 5/36 | 15/36 | 45/36 |
4 | 7/36 | 28/36 | 112/36 |
5 | 9/36 | 45/36 | 225/36 |
6 | 11/36 | 66/36 | 396/36 |
Total | 161/36 | 791/36 |
So,
--------------------------------------------------------
Following table shows the calculations for mean and var of V:
V | P(V=v) | v*P(V=v) | v^2*P(V=v) |
1 | 11/36 | 11/36 | 11/36 |
2 | 9/36 | 18/36 | 36/36 |
3 | 7/36 | 21/36 | 63/36 |
4 | 5/36 | 20/36 | 80/36 |
5 | 3/36 | 15/36 | 75/36 |
6 | 1/36 | 6/36 | 36/36 |
Total | 91/36 | 301/36 |
So,
The covariance is
Cov(U,V) = E(UV) - E(U) E(V) = 12.25 - (161/36) * (91/36) = 0.9452
Therefore correlation coefficient is