Summation Examples
There are times when an infinite series must be evaluated. Some of these problems are easy to solve and others are more challenging.
1. [Exercise 2.3.4(e)] Let X equal the number of rolls of a pair of dice that are needed to determine whether a bettor wins or loses playing craps. The p.d.f. of X is f (1) = 12/36 and
f ( x ) = 2[(3/36)(9/36)(27/36)^( x -2) + (4/36)(10/36)(26/36)^( x -2) + (5/36)(11/36)(25/36)^( x -2)]
or
for x = 2, 3, ... . Show that
E ( X ) = m = 557/165
and
Var(
X
) =
= 245,672/27,255.
2. [Exercise 3.6.11] Flip n fair coins until heads has been observed on each coin. E.g. toss the coins and remove the heads, toss the remaining coins and remove the heads, etc. Let X equal the number of tosses required. It can be shown that the p.d.f. of X is
.
Find m = E ( X ). How does the value of n affect m ? Solution
3. Flip a coin successively. Let X equal the number of flips needed to observe
(a) [Exercises 1.1.6 and 3.6.20] The same face on successive flips.
(b) [Exercises 1.1.7 and 3.6.21] HH on successive flips.
(c) [Exercises 1.1.8 and 3.6.22] HT on successive flips.
4. Let
X
have a Poisson distribution with mean
m
=
.
(a) [Exercise 3.7.5] Find the mean, variance, and other moments of X symbolically.
(b) [Exercise 6.1.1] Both
X
-bar and the sample variance,
, are unbiased estimators for
l
. Compare the variances of these two estimators both theoretically and empirically.
Solution