This page intentionally left blank

268 CHAPTER 6 The Normal Distribution and Other Continuous Distributions

EXPONENTIAL PROBABILITY DENSITY FUNCTION

f 1X2 = le-lx for X 7 0 (6.7)
where
e = mathematical constant approximated by 2.71828
l = mean number of arrivals per unit
X = any value of the continuous variable where 0 6 X 6 q

The mean time between arrivals, m, is given by Equation (6.8).

MEAN TIME BETWEEN ARRIVALS

1
m = (6.8)
l

The standard deviation of the time between arrivals, s, is given by Equation (6.9).

STANDARD DEVIATION OF THE TIME BETWEEN ARRIVALS

1
s = (6.9)
l

The value 1>l is equal to the mean time between arrivals. For example, if the mean num-
ber of arrivals in a minute is l = 4, then the mean time between arrivals is 1>l = 0.25 min-
utes, or 15 seconds. Equation (6.10) defines the cumulative probability that the length of time
before the next arrival is less than or equal to X.

CUMULATIVE EXPONENTIAL PROBABILITY

P1arrival time … X 2 = 1 - e-lx (6.10)

To illustrate the exponential distribution, suppose that customers arrive at a bank’s ATM at
a rate of 20 per hour. If a customer has just arrived, what is the probability that the next
customer will arrive within 6 minutes (i.e., 0.1 hour)? For this example, l = 20 and X = 0.1.
Using Equation (6.10),
P1Arrival time … 0.12 = 1 - e-2010.12
= 1 - e-2
= 1 - 0.1353 = 0.8647

Thus, the probability that a customer will arrive within 6 minutes is 0.8647, or 86.47%. Fig-
ure 6.25 shows this probability as computed by Excel (left results) and Minitab (right results).
 Problems for Section 6.5 269

FIGURE 6.25
Excel and Minitab results
for finding exponential
probabilities
1mean = 1>l2

EXAMPLE 6.6 In the ATM example, what is the probability that the next customer will arrive within 3 min-
utes (i.e., 0.05 hour)?
Computing
Exponential SOLUTION For this example, l = 20 and X = 0.05. Using Equation (6.10),
Probabilities P1Arrival time … 0.052 = 1 - e-2010.052
= 1 - e-1
= 1 - 0.3679 = 0.6321
Thus, the probability that a customer will arrive within 3 minutes is 0.6321, or 63.21%.

Problems for Section 6.5

LEARNING THE BASICS c. What are your answers to (a) and (b) if the rate of arrival
6.28 Given an exponential distribution with l = 10, what of autos is 60 per minute?
is the probability that the arrival time is d. What are your answers to (a) and (b) if the rate of arrival
a. less than X = 0.1? of autos is 30 per minute?
b. greater than X = 0.1? SELF 6.32 Customers arrive at the drive-up window of
c. between X = 0.1 and X = 0.2? Test a fast-food restaurant at a rate of 2 per minute dur-
d. less than X = 0.1 or greater than X = 0.2? ing the lunch hour.
6.29 Given an exponential distribution with l = 30, what a. What is the probability that the next customer will arrive
is the probability that the arrival time is within 1 minute?
a. less than X = 0.1? b. What is the probability that the next customer will arrive
b. greater than X = 0.1? within 5 minutes?
c. between X = 0.1 and X = 0.2? c. During the dinner time period, the arrival rate is 1 per
d. less than X = 0.1 or greater than X = 0.2? minute. What are your answers to (a) and (b) for this period?

6.30 Given an exponential distribution with l = 5, what is 6.33 Telephone calls arrive at the information desk of a
the probability that the arrival time is large computer software company at a rate of 15 per hour.
a. less than X = 0.3? a. What is the probability that the next call will arrive
b. greater than X = 0.3? within 3 minutes (0.05 hour)?
c. between X = 0.3 and X = 0.5? b. What is the probability that the next call will arrive
d. less than X = 0.3 or greater than X = 0.5? within 15 minutes (0.25 hour)?
c. Suppose the company has just introduced an updated ver-
sion of one of its software programs, and telephone calls
APPLYING THE CONCEPTS are now arriving at a rate of 25 per hour. Given this infor-
mation, what are your answers to (a) and (b)?
6.31 Autos arrive at a toll plaza located at the entrance to a
bridge at a rate of 50 per minute during the 5:00-to-6:00 P.M. 6.34 An on-the-job injury occurs once every 10 days on
hour. If an auto has just arrived, average at an automobile plant. What is the probability that
a. what is the probability that the next auto will arrive the next on-the-job injury will occur within
within 3 seconds (0.05 minute)? a. 10 days?
b. what is the probability that the next auto will arrive b. 5 days?
within 1 second (0.0167 minute)? c. 1 day?
270 CHAPTER 6 The Normal Distribution and Other Continuous Distributions

6.35 The time between unplanned shutdowns of a power c. The actual arrival rate on Fridays is 15 per hour. What are
plant has an exponential distribution with a mean of your answers to (a) and (b) for Fridays?
20 days. Find the probability that the time between two
6.37 Some Internet companies sell a service that will boost
unplanned shutdowns is
a website’s traffic by delivering additional unique visitors.
a. less than 14 days.
Assume that one such company claims it can deliver
b. more than 21 days.
1,000 visitors a day. If this amount of website traffic is ex-
c. less than 7 days.
perienced, then the time between visitors has a mean of
6.36 Golfers arrive at the starter’s booth of a public golf 1.44 minutes (or 0.6944 per minute). Assume that your
course at a rate of 8 per hour during the Monday-to-Friday website gets 1,000 visitors a day and that the time between
midweek period. If a golfer has just arrived, visitors has an exponential distribution. What is the proba-
a. what is the probability that the next golfer will arrive bility that the time between two visitors is
within 15 minutes (0.25 hour)? a. less than 1 minute?
b. what is the probability that the next golfer will arrive b. less than 2 minutes?
within 3 minutes (0.05 hour)? c. more than 3 minutes?
d. Do you think it is reasonable to assume that the time
between visitors has an exponential distribution?

6.6 Online Topic: The Normal Approximation

to the Binomial Distribution
In many circumstances, you can use the normal distribution to approximate the binomial dis-
tribution. To study this topic, read the Section 6.6 online topic file that is available on this
book’s companion website. (See Appendix C to learn how to access the online topic files.)

U S I N G S TAT I S T I C S @ OurCampus! Revisited

I n the OurCampus! scenario, you were a designer for a social net-

working website. You sought to ensure that a video could be down-
loaded quickly for playback in the web browsers of site visitors.
(Quick playback of videos would help attract and retain those visi-
tors.) By running experiments in the corporate offices, you determined that
the amount of time, in seconds, that passes from first linking to the website
until a video is fully displayed is a bell-shaped distribution with a mean
download time of 7 seconds and standard deviation of 2 seconds. Using the
normal distribution, you were able to calculate that approximately 84% of the download times
are 9 seconds or less, and 95% of the download times are between 3.08 and 10.92 seconds.
Now that you understand how to calculate probabilities from the normal distribution, you
can evaluate download times of a video using different web page designs. For example, if the
standard deviation remained at 2 seconds, lowering the mean to 6 seconds would shift the en-
tire distribution lower by 1 second. Thus, approximately 84% of the download times would be
8 seconds or less, and 95% of the download times would be between 2.08 and 9.92 seconds.
Another change that could reduce long download times would be reducing the variation. For
example, consider the case where the mean remained at the original 7 seconds but the standard
deviation was reduced to 1 second. Again, approximately 84% of the download times would
be 8 seconds or less, and 95% of the download times would be between 5.04 and 8.96 seconds.

SUMMARY
In this and the previous chapter, you have learned about order forms in a report generated by an accounting informa-
mathematical models called probability distributions and tion system). In this chapter, you learned about continuous
how they can be used to solve business problems. In Chapter probability distributions where the outcomes come from a
5, you used discrete probability distributions in situations measuring process (e.g., your height, the download time of a
where the outcomes come from a counting process (e.g., the video). Continuous probability distributions come in various
number of courses you are enrolled in, the number of tagged shapes, but the most common and most important in business
 Chapter Review Problems 271

is the normal distribution. The normal distribution is sym- using the normal distribution. In Section 6.3, you learned
metrical; thus, its mean and median are equal. It is also bell- about various methods for evaluating normality in order to
shaped, and approximately 68.26% of its observations are determine whether the normal distribution is a reasonable
within 1 standard deviation of the mean, approximately mathematical model to use in specific situations. In Sections
95.44% of its observations are within 2 standard deviations 6.4 and 6.5, you studied continuous distributions that were
of the mean, and approximately 99.73% of its observations not normally distributed—in particular, the uniform and ex-
are within 3 standard deviations of the mean. Although many ponential distributions.
data sets in business are closely approximated by the normal Chapter 7 uses the normal distribution to develop the
distribution, do not think that all data can be approximated subject of statistical inference.

K E Y E Q U AT I O N S
Normal Probability Density Function Variance and Standard Deviation of the Uniform
Distribution
f 1X 2 =
1
e-11>2231X - m2>s4
2

1b - a22
(6.1)
22ps s2 = (6.6a)
12
1b - a22
Transformation Formula
s = (6.6b)
X - m A 12
Z = (6.2)
s
Exponential Probability Density Function
f 1X 2 = le-lx for X 7 0
Finding an X Value Associated with a Known Probability
(6.7)
X = m + Zs (6.3)
Mean Time Between Arrivals
Uniform Probability Density Function
1
1 m = (6.8)
f1X2 = (6.4) l
b - a
Standard Deviation of the Time Between Arrivals
Mean of the Uniform Distribution
1
a + b s = (6.9)
m = (6.5) l
2
Cumulative Exponential Probability
P1arrival time … X 2 = 1 - e-lx (6.10)

KEY TERMS
cumulative standardized normal probability density function 248 standardized normal random
distribution 251 probability density function for the variable 250
exponential distribution 267 normal distribution 250 transformation formula 250
normal distribution 248 quantile–quantile plot 262 uniform distribution 265
normal probability plot 262 rectangular distribution 265

CHAPTER REVIEW PROBLEMS

CHECKING YOUR UNDERSTANDING 6.39 How do you find the area between two values under
6.38 Why is only one normal distribution table such as the normal curve?
Table E.2 needed to find any probability under the normal 6.40 How do you find the X value that corresponds to a
curve? given percentile of the normal distribution?
272 CHAPTER 6 The Normal Distribution and Other Continuous Distributions

6.41 What are some of the distinguishing properties of a d. 80% of the oranges contain between what two values (in
normal distribution? ounces of juice), symmetrically distributed around the
population mean?
6.42 How does the shape of the normal distribution differ
from the shapes of the uniform and exponential distributions? 6.49 The file DomesticBeer contains the percentage alco-
hol, number of calories per 12 ounces, and number of car-
6.43 How can you use the normal probability plot to eval-
bohydrates (in grams) per 12 ounces for 139 of the
uate whether a set of data is normally distributed?
best-selling domestic beers in the United States. For each of
6.44 Under what circumstances can you use the exponen- the three variables, decide whether the data appear to be ap-
tial distribution? proximately normally distributed. Support your decision
through the use of appropriate statistics and graphs.
APPLYING THE CONCEPTS Source: Data extracted from www.Beer100.com, March 18, 2010.
6.45 An industrial sewing machine uses ball bearings that
are targeted to have a diameter of 0.75 inch. The lower and 6.50 The evening manager of a restaurant was very
upper specification limits under which the ball bearings can concerned about the length of time some customers were
operate are 0.74 inch and 0.76 inch, respectively. Past expe- waiting in line to be seated. She also had some concern
rience has indicated that the actual diameter of the ball bear- about the seating times—that is, the length of time between
ings is approximately normally distributed, with a mean of when a customer is seated and the time he or she leaves the
0.753 inch and a standard deviation of 0.004 inch. What is restaurant. Over the course of one week, 100 customers (no
the probability that a ball bearing is more than 1 per party) were randomly selected, and their
a. between the target and the actual mean? waiting and seating times (in minutes) were recorded in
Wait .
b. between the lower specification limit and the target?
c. above the upper specification limit? a. Think about your favorite restaurant. Do you think wait-
d. below the lower specification limit? ing times more closely resemble a uniform, an exponen-
e. Of all the ball bearings, 93% of the diameters are greater tial, or a normal distribution?
than what value? b. Again, think about your favorite restaurant. Do you think
seating times more closely resemble a uniform, an expo-
6.46 The fill amount in 2-liter soft drink bottles is nor- nential, or a normal distribution?
mally distributed, with a mean of 2.0 liters and a standard c. Construct a histogram and a normal probability plot of
deviation of 0.05 liter. If bottles contain less than 95% of the the waiting times. Do you think these waiting times more
listed net content (1.90 liters, in this case), the manufacturer closely resemble a uniform, an exponential, or a normal
may be subject to penalty by the state office of consumer distribution?
affairs. Bottles that have a net content above 2.10 liters may d. Construct a histogram and a normal probability plot of
cause excess spillage upon opening. What proportion of the the seating times. Do you think these seating times more
bottles will contain closely resemble a uniform, an exponential, or a normal
a. between 1.90 and 2.0 liters? distribution?
b. between 1.90 and 2.10 liters?
c. below 1.90 liters or above 2.10 liters? 6.51 All the major stock market indexes posted strong gains
d. At least how much soft drink is contained in 99% of the in 2009. The mean one-year return for stocks in the S&P 500,
bottles? a group of 500 very large companies, was 23.45%. The mean
e. 99% of the bottles contain an amount that is between which one-year return for the NASDAQ, a group of 3,200 small and
two values (symmetrically distributed) around the mean? medium-sized companies, was 43.89%. Historically, the one-
year returns are approximately normally distributed, the stan-
6.47 In an effort to reduce the number of bottles that contain dard deviation in the S&P 500 is approximately 20%, and the
less than 1.90 liters, the bottler in Problem 6.46 sets the filling standard deviation in the NASDAQ is approximately 30%.
machine so that the mean is 2.02 liters. Under these circum- a. What is the probability that a stock in the S&P 500
stances, what are your answers in Problem 6.46 (a) through (e)? gained value in 2009?
6.48 An orange juice producer buys all his oranges from a b. What is the probability that a stock in the S&P 500
large orange grove. The amount of juice squeezed from each gained 10% or more?
of these oranges is approximately normally distributed, with a c. What is the probability that a stock in the S&P 500 lost
mean of 4.70 ounces and a standard deviation of 0.40 ounce. 20% or more in 2009?
a. What is the probability that a randomly selected orange d. What is the probability that a stock in the S&P 500 lost
will contain between 4.70 and 5.00 ounces of juice? 40% or more?
b. What is the probability that a randomly selected orange e. Repeat (a) through (d) for a stock in the NASDAQ.
will contain between 5.00 and 5.50 ounces of juice? f. Write a short summary on your findings. Be sure to in-
c. At least how many ounces of juice will 77% of the clude a discussion of the risks associated with a large
oranges contain? standard deviation.