Problems 609
and make the change of variables
x=αtβ, dx=αβtβ−^1 dt
24.Show that ifXis a Weibull random variable with parameters (α,β), then
Var(X)=α−2/β
[
(
1 +
2
β
)
−
(
(
1 +
1
β
)) 2 ]
25.If the following are the sample data from a Weibull population having unknown
parametersαandβ, determine the least square estimates of these quantities, using
either of the methods presented.
Data:15.4, 16.8, 6.2, 10.6, 21.4, 18.2, 1.6, 12.5, 19.4, 17
26.Show that ifXis a Weibull random variable with parameters (α,β), thenαXβis
an exponential random variable with mean 1.
27.IfU is uniformly distributed on (0, 1) — that is,Uis a random number —
show that [−(1/α) logU]1/β is a Weibull random variable with parameters
(α,β).
The next three problems are concerned with verifying Equations 14.5.5 and
14.5.7.
28.IfXis a continuous random variable having distribution functionF, show that
(a) F(X) is uniformly distributed on (0, 1);
(b) 1 −F(X) is uniformly distributed on (0, 1).
29.LetX(i)denoteith smallest of a sample of sizenfrom a continuous distribution
functionF. Also, letU(i)denote theith smallest from a sample of sizenfrom a
uniform (0, 1) distribution.
(a) Argue that the density function ofU(i)is given by
fU(i)(t)=
n!
(n−i)!(i−1)!
ti−^1 (1−t)n−i,0<t< 1
[Hint:In order for theith smallest ofnuniform (0, 1) random variables to
equalt, how many must be less thantand how many must be greater? Also,
in how many ways can a set ofnelements be broken into three subsets of
respective sizesi−1, 1, andn−i?]