c03RandomVariablesandProbabilityDistributions.

qxd 9/21/10 9:34 AM Page 72

72 CHAPTER 3 RANDOM VARIABLES AND PROBABILITY DISTRIBUTIONS

EXAMPLE 3-5 For the copper current measurement in Example 3-2, the mean of X is
Current in a Wire:
20
Mean
冮 x f (x) dx  冮 x a 20 b dx  0.05 x 2 `
20
1
E(X )  2
 10
0
 0

The variance of X is

20

冮 (x  ) f (x) dx  冮 (x  10) a 20 b dx  0.05(x  10) 3 `

20
1
V(X )  2 2 3
 33.33 ■
0
 0

EXAMPLE 3-6 For the distance to a flaw in Example 3-2, the mean of X is
Current in a Wire:
Variance
ex 2000
E(X )  冮 x f (x) dx  冮 x 2000
dx
0 0

A table of integrals or integration by parts can be used to show that

E(X )  xex  2000 ` 

0
冮e x  2000
dx  0  2000 ex  2000 `  2000
0
0

The variance of X is

ex  2000
V(X )  冮 (x  ) f (x) dx  冮 (x  2000)
2 2
2000
dx
 0

A table of integrals or integration by parts can be applied twice to show that

V(X )  2000 2  4,000,000 ■

EXERCISES FOR SECTION 3-4

3-21. Show that the following functions are probability den- (a) P(X  6) (b) P(6  X 8)
sity functions for some value of k and determine k. Then (c) P(X 8) (d) P(X  8)
determine the mean and variance of X. (e) Determine x such that P(X x)  0.95.
(a) f (x)  kx 2 for 0 x 4 3-24. Suppose that f (x)  1.5x 2 for 1 x 1 and f (x)  0
(b) f (x)  k(1  2x) for 0 x 2 otherwise. Determine the following probabilities.
(c) f (x)  kex for 0 x (a) P(0 X ) (b) P(0.5 X )
(d) f (x)  k where k  0 and 100 x 100  k (c) P(0.5  X  0.5) (d) P(X 2)
3-22. For each of the density functions in Exercise 3-21, (e) P(X 0 or X  0.5)
perform the following. (f ) Determine x such that P(x X )  0.05.
(a) Graph the density function and mark the location of the 3-25. The pdf of the time to failure of an electronic compo-
mean on the graph. nent in a copier (in hours) is f (x)  [exp (x 兾3000)] 兾3000
(b) Find the cumulative distribution function. for x  0 and f (x)  0 for x  0. Determine the probability that
(c) Graph the cumulative distribution function. (a) A component lasts more than 1000 hours before failure.
3-23. Suppose that f (x)  e(x6) for 6 x and f (x)  0 for (b) A component fails in the interval from 1000 to 2000
x  6. Determine the following probabilities. hours.
[Link] 9/21/10 9:34 AM Page 73

3-4 CONTINUOUS RANDOM VARIABLES 73

(c) A component fails before 3000 hours. 7 micrometers in size. What proportion of the particles
(d) Determine the number of hours at which 10% of all com- will be detected?
ponents have failed. 3-30. (Integration by parts is required in this exercise.) The
(e) Determine the mean. probability density function for the diameter of a drilled hole
3-26. The probability density function of the net weight in in millimeters is 10e10(x5) for x  5 mm and zero for x  5
ounces of a packaged compound is f (x)  2.0 for 19.75 x mm. Although the target diameter is 5 millimeters, vibrations,
20.25 ounces and f (x)  0 for x elsewhere. tool wear, and other factors can produce diameters larger than
(a) Determine the probability that a package weighs less than 5 millimeters.
20 ounces. (a) Determine the mean and variance of the diameter of the
(b) Suppose that the packaging specifications require that the holes.
weight be between 19.9 and 20.1 ounces. What is the (b) Determine the probability that a diameter exceeds 5.1 mil-
probability that a randomly selected package will have a limeters.
weight within these specifications? 3-31. Suppose the cumulative distribution function of the
(c) Determine the mean and variance. length (in millimeters) of computer cables is
(d) Find and graph the cumulative distribution function.
3-27. The temperature readings from a thermocouple in a 0 x  1200
furnace fluctuate according to a cumulative distribution F(x)  • 0.1x  120 1200 6 x  1210
function 1 x 7 1210

0 x 6 800°C (a) Determine P(x 1208).

F(x)  • 0.1x  80 800°C  x 6 810°C (b) If the length specifications are 1195 x 1205 millime-
1 x 7 810°C ters, what is the probability that a randomly selected com-
puter cable will meet the specification requirement?
Determine the following. 3-32. The thickness of a conductive coating in micrometers
(a) P(X 805) (b) P(800 X  805) (c) P(X  808) has a density function of 600x2 for 100 m x 120 m
(d) If the specifications for the process require that the fur- and zero for x elsewhere.
nace temperature be between 802 and 808C, what is the (a) Determine the mean and variance of the coating thick-
probability that the furnace will operate outside of the ness.
specifications? (b) If the coating costs $0.50 per micrometer of thickness on
3-28. The thickness measurement of a wall of plastic tub- each part, what is the average cost of the coating per part?
ing, in millimeters, varies according to a cumulative distribu- 3-33. A medical linear accelerator is used to accelerate elec-
tion function trons to create high-energy beams that can destroy tumors
with minimal impact on surrounding healthy tissue. The beam
0 x 6 2.0050 energy fluctuates between 200 and 210 MeV (million electron
F(x)  • 200x  401 2.0050  x  2.0100 volts). The cumulative distribution function is
1 x 7 2.0100
0 x 6 200
Determine the following. F(x)  • 0.1x  20 200  x  210
(a) P(X  2.0080) (b) P(X  2.0055) 1 x 7 210
(c) If the specification for the tubing requires that the thick-
ness measurement be between 2.0090 and 2.0100 mil- Determine the following.
limeters, what is the probability that a single measurement (a) P(X 209) (b) P(200 X 208) (c) P(X  209)
will indicate conformance to the specification? (d) What is the probability density function?
3-29. Suppose that contamination particle size (in micro- (e) Graph the probability density function and the cumulative
meters) can be modeled as f (x)  2x3 for 1 x and f (x)  0 distribution function.
for x  1. (f ) Determine the mean and variance of the beam energy.
(a) Confirm that f(x) is a probability density function. 3-34. The probability density function of the time a cus-
(b) Give the cumulative distribution function. tomer arrives at a terminal (in minutes after 8:00 A.M.) is
(c) Determine the mean. f (x)  0.1ex 10 for x 7 0. Determine the probability that
(d) What is the probability that the size of a random particle (a) The customer arrives by 9:00 A.M.
will be less than 5 micrometers? (b) The customer arrives between 8:15 A.M. and 8:30 A.M.
(e) An optical device is being marketed to detect contamina- (c) Determine the time at which the probability of an earlier
tion particles. It is capable of detecting particles exceeding arrival is 0.5.
[Link] 11/10/10 5:17 PM Page 74

74 CHAPTER 3 RANDOM VARIABLES AND PROBABILITY DISTRIBUTIONS

(d) Determine the cumulative distribution function and use (a) P(X 6 1)
the cumulative distribution function to determine the (b) P(X 7 2)
probability that the customer arrives between 8:15 A.M. (c) P(1  X 6 2)
and 8:30 A.M. (d) x such that P(X 6 x)  0.95
(e) Determine the mean and standard deviation of the number (e) pdf
of minutes until the customer arrives. 3-37. The waiting time until service at a hospital emergency
3-35. The probability density function of the weight of department is modeled with the pdf f (x)  (1兾9)x for
packages delivered by a post office is f (x)  70(69x2 ) for 0 6 x 6 3 hours and f (x)  2 3  (19) x for 3 6 x 6 6.
1 6 x 6 70 pounds. Determine the following:
(a) What is the probability a package weighs less than 10 (a) Probability the wait is less than 4 hours
pounds? (b) Probability the wait is more than 5 hours
(b) Determine the mean and variance of package weight. (c) Probability the wait is less than or equal to 30 minutes
(c) If the shipping cost is $3 per pound, what is the average (d) Waiting time that is exceeded by only 10% of patients
shipping cost of a package? (e) Mean waiting time
3-36. Given the cdf F(x)  0 for x 6 0, 1  exp (x2) for
0 6 x, determine the following:

3-5 IMPORTANT CONTINUOUS DISTRIBUTIONS

3-5.1 Normal Distribution

Undoubtedly, the most widely used model for the distribution of a random variable is a normal
distribution. In Chapter 2, several histograms are shown with similar symmetric, bell shapes.
A fundamental result, known as the central limit theorem, implies that histograms often have
this characteristic shape, at least approximately. Whenever a random experiment is replicated,
the random variable that equals the average (or total) result over the replicates tends to have
a normal distribution as the number of replicates becomes large. De Moivre presented initial
results in 1733 and Gauss developed a normal distribution nearly 100 years later. Consequently,
a normal distribution is also referred to as a Gaussian distribution.
When do we average (or total) results? Almost always. In Example 2-1 the average of the
eight tensile strength measurements was calculated to be 1055.0 psi. If we assume that each
measurement results from a replicate of a random experiment, the normal distribution can be
used to make approximate conclusions about this average. Such conclusions are the primary
topics in the subsequent chapters of this book.
Furthermore, sometimes the central limit theorem is less obvious. For example, assume
that the deviation (or error) in the length of a machined part is the sum of a large number of
infinitesimal (small) effects, such as temperature and humidity drifts, vibrations, cutting angle
variations, cutting tool wear, bearing wear, rotational speed variations, mounting and fixturing
variations, variations in numerous raw material characteristics, and variation in levels of con-
tamination. If the component errors are independent and equally likely to be positive or nega-
tive, the total error can be shown to have an approximate normal distribution. Furthermore, the
normal distribution arises in the study of numerous basic physical phenomena. For example,
the physicist Maxwell developed a normal distribution from simple assumptions regarding the
velocities of molecules.
The theoretical basis of a normal distribution is mentioned to justify the somewhat com-
plex form of the probability density function. Our objective now is to calculate probabilities
for a normal random variable. The central limit theorem will be stated more carefully later in
this chapter.