8.

1 Population Mean: Known σ 305

Therefore, for 90% confidence, the margin of error is 3.29 and the confidence interval is 82.Ϫ
3.29ϭ 78.71 to 82ϩ 3.29ϭ 85.29. Similarly, the 99% confidence interval is
20
82 Ϯ 2.576
͙ 100
82 Ϯ 5.15
So, for 99% confidence, the margin of error is 5.15 and the confidence interval is 82.Ϫ
5.15ϭ 76.85 to 82ϩ 5.15ϭ 87.15
When comparing the results for the 90%, 95%, and 99% confidence levels, it is clear that
to have a greater degree of confidence, the margin of error and with this the width of the interval of
trust must be greater.

Practical recommendation
If the population has a normal distribution, the confidence interval obtained with the ex-
pressure (8.1) is exact. In other words, if expression (8.1) is used repeatedly to generate.
95% confidence intervals, exactly 95% of the generated intervals will contain the
population mean. If the population does not have a normal distribution, the confidence interval
obtained with expression (8.1) will be approximate. In this case, the quality of the approximation of-
It depends both on the distribution of the population and on the size of the sample.
In most applications, when the expression (8.1) is used, a sample size
nՆ 30 is suitable for obtaining an interval estimate of the population mean. If the popu-
the distribution is not normally distributed, but it is more or less symmetric, sample sizes of up to
15 can be expected to provide a good approximation of the confidence interval. With th-
smaller sample sizes, expression (8.1) should only be used if the analyst believes, or is willing to
It is assumed that the distribution of the population is at least approximately normal.

NOTES AND COMMENTS

1.The interval estimation procedure vis- 2.The size of the sample appears in the denominator
this section is based on the assumption that Nador of the expression (8.1) for the estimation
the population standard deviation σ is known by interval. Consequently, if one determines-
Yes. To say that it is known means that it is counted. of sample size gives an interval too much-
has historical data or with other information that to be broad, so that it has practical usefulness, it
allow to obtain a good estimate of the des- increase the sample size. Synesthesia in the
Standard procedure before taking the sample that is denominator, with a larger sample size
will be used to obtain the estimate of the mean a smaller margin of error will be obtained, an inter-
blational. So technically this does not narrower and greater precision. The proce-
It means that it is known for sure. What measurement to determine the size of a
it means that only a good estimate was obtained simple random sample needed to obtain
calculation of the standard deviation before taking the having a certain precision will be seen in the
show and that in this way the misuse will not be used section 8.3.
my sample to estimate both the population mean
as the population standard deviation.

Exercises

Methods
In a simple random sample of 40 items, the sample mean obtained is 25. The standard deviation
population standard is σϭ 5.
a. What is the standard error of the mean σx̄ ?
b. What is the margin of error to have 95% confidence?
306 Chapter 8 Interval Estimation

In a simple random sample of 50 items from a population in which σϭ 6 the average sample
Carexam tral was 32.
a. Provide a 90% confidence interval for the population mean.
b. Establish a 95% confidence interval for the population mean.
c. Provide a 99% confidence interval for the population mean.
In a simple random sample of 60 items, the sample mean was 80. The standard deviation
population isσϭ 15.
a. Calculate the 95% confidence interval for the population mean.
b. Suppose that the same sample mean was obtained from a sample of 120 items. Provide the
95% confidence interval for the population mean.
4. For the population mean, the following 95% confidence interval was given, from 152 to 160. If
σ ϭ 15, what is the sample size used in this study?

Applications
5. In order to estimate the average amount spent by a customer on a meal at an important
Carexam restaurant, data was collected from a sample of 49 customers. Assume that the deviation is-
the population standard is $5.
a. What is the margin of error for 95% confidence?
b. If the population mean is $24.80, what is the 95% confidence interval for the mean?
blational?

file CD
6. Nielsen Media Research conducted a study to find out how much time was spent watching television.
the homes, from 8:00 to 11:00 at night. The data found in the file
in Nielsen of the compact disc is consistent with the reported findings (The World Almanac,
Nielsen 2003). Based on previous studies, the population standard deviation is considered known.
the day isϭ 3.5 hours. Provide an estimate using a 95% confidence interval for the me-
day of the time spent watching television per week from 8:00 to 11:00 in the evening.
In a study about small businesses that have a website, it was found that
the average amount spent at a site is $11,500 per year. Given a sample of 60 business
means and a standard deviation σϭ $4000, what is the margin of error? Use 95% confidence.
What would you recommend if the study requires a margin of error of $500?
8. The National Quality Research Center at the University of Michigan provides quarterly measures
trails of consumer opinions about products and services (The Wall Street Jour-
nal, February 18, 2003). In a survey about 10 fast food and pizza restaurants, the
The average consumer satisfaction index was 71. Previous data indicates that the des-
Standard aviation has been relatively stable and isϭ 5.
a. What must the researcher be willing to assume if they want a margin of error?
b. With 95% confidence, what is the margin of error?
c. What is the margin of error if a 99% confidence is desired?
9. The average score on the admission exam of the students who entered the school of
business was 3.37 (Best Graduate Schools, U.S. News and World Report, 2001). Assume that I-
This estimate was based on a sample of 120 students. According to previous data, it was indicated...
what is known as the standard deviation and what is σϭ 0.28. What is the estimate using a
95% confidence interval for the average score of the students entering-
Do they go to business school?
10. The magazine Playbill reported that the average annual household income of its subscribers is $119,155.
(Playbill, January 2006). Suppose that the estimate of the average annual household income is based on
from a sample of 80 families and based on data from previous studies, the standard deviation po-
blacional is known and isϭ $30 000.
 8.2 Population mean: σunknown 307

a. Develop a 90% confidence interval for the population mean.

b. Give a 95% confidence interval estimate for the population mean.
c. Give a 99% confidence interval estimation for the population mean.
d. What happens to the amplitude of the confidence interval as the confidence level increases?
mint? Does this seem reasonable? Explain.

8.2Population mean: unknown standard deviation

When calculating a confidence interval for the population mean, it is often the case that one does not have a
good estimate of the population standard deviation. In such cases, the same sample is used.
to estimate μyσ. This situation is the case known asunknown.When to use
To estimate σ, the margin of error, and the interval estimate of the population mean is based on
in a probability distribution known asdistribution.Although the preparation
mathematics of distribution part of the assumption that the population being sampled
William Sealy Gosset, who
published under the
It has a normal distribution, research has shown that the distribution applies.
pseudonym 'Student' in many situations where the population significantly deviates from a normal population.
established the t distribution. Later in this same section, guidelines for using the distribution are provided.
Gosset, who had when the population is not normally distributed.
studied mathematics in
Oxford, was working for
The distribution is a family of similar probability distributions; each distribution
Guinness Brewery in it depends on a parameter known asdegrees of freedom. The distribution for a degree
Dublin, Ireland. It elaborated the The distribution of freedom is unique, as it is for two degrees of freedom, for three degrees of
t distribution when freedom, etc. As the number of degrees of freedom increases, the difference between the distribution-
I was working with materials at
small scale and towards
the standard normal distribution is decreasing. Figure 8.4 shows the dis-
experiments of distributions for 10 and 20 degrees of freedom and their relationship with the probability distribution
temperature. standard normal. Note that a distribution for more degrees of freedom exhibits less variance.

FIGURE 8.4COMPARISON OF THE STANDARD NORMAL DISTRIBUTION WITH

THE DISTRIBUTIONS FOR 10 AND 20 DEGREES OF FREEDOM

Standard normal distribution

Distribution (20 degrees of freedom)

Distribution (10 degrees of freedom)

z,t
0
314 Chapter 8 Interval Estimation

NOTES AND COMMENTS

1. When the margin of error is known, greater values ofxb

-aIrn this way, when
zα/2(σ͞ ͙ n) , it is fixed and it is the same for all the isx̅ greater than the population average, tends to
size samples. When it is not known, to be greater than σ. This bias causes the margin
margin of errortα/2(s͞ n ) ,͙ varies from a sample- of error,tα/2(s͙͞ n ) is greater than it would be
to another. This variation is due to the deviation- if it were knownσA confidence interval
The standard sample deviation varies with the sample. with a larger margin of error it will tend to include
if it is large, a sea is obtained more frequently to the population mean μ
big error, while small ones that if the true value were usedσ.But how-
then, a small margin of error is obtained. doisx̄ below the population average, the co-
What happens with interval estimates? relationship betweenx
basrince the margin of
When is the population biased? Consider a error is smaller. In this case, such intervals
right-skewed population, in which the trustworthy with a smaller margin of error include-
data with large values pull the distribution they were to the average population fewer times than if
to the right. When there is such a bias, it would be known and used. For this reason, it re-
there is a positive correlation between the mean it is advisable to use larger sample sizes
x̄ the sample standard deviation.
sampleand since when the distribution of the population is
Highervaluestendtocorrespondto very biased.

Exercises

Methods
11. In the distribution with 16 degrees of freedom, find the area, or the probability, of each one.
from the following regions:
to the right of 2.120
b. To the left of 1.337
to the left ofϪ1.746
to the right of 2.583
and amongϪ2.120 and 2.120
f. BetweenϪ1.746 and 1.746
12. Find the values for the following situations.
An area of 0.025 in the upper tail, with 12 degrees of freedom
b. An area of 0.05 in the lower tail, with 50 degrees of freedom
c. An area of 0.01 in the upper tail, with 30 degrees of freedom
d. Among those that remain 90% of the area, with 25 degrees of freedom
e. Among those remaining 95% of the area, with 45 degrees of freedom
13. The following sample data comes from a normal population: 10, 8, 12, 15, 13, 11, 6, 5.
Carexam a. What is the point estimate of the population mean?
b. What is the point estimate of the population standard deviation?
c. With 95% confidence, what is the margin of error for the estimation of the mean?
blacional?
d. What is the 95% confidence interval for the population mean?
14. In a simple random sample withϭ The sample mean was 22.5 and the standard deviation
sample 4.4.
a. Find a 90% confidence interval for the population mean.
b. Give a 95% confidence interval for the population mean.
Give a 99% confidence interval for the population mean.
What happens to the margin of error and the confidence interval as it increases?
What's the confidence level?
 8.2 Population mean: unknown σ 315

Applications
15. The sales agents of a company present a weekly report that lists the clients.
Carexam contacted during the week. In a sample of 65 weekly reports, the sample mean is
19.5 clients per week. The standard deviation is 5.2. Provide 90% confidence intervals.
95% for the population mean of the number of clients contacted weekly by the staff
sales branch.
The average number of flight hours for Continental Airlines pilots is 49 hours per month.
(The Wall Street Journal, February 25, 2003). Assume that this average was based on the hours of
flight of a sample of 100 pilots from that company and that the sample standard deviation has
I sat for 8.5 hours.
a. At 95% confidence, what is the margin of error?
b. Give the 95% confidence interval for the population mean of flight hours of the
pilots.
c. The average flight hours of United Airlines pilots is 36 hours per month. Use
the results of item b to analyze the difference in the number of flight hours of
the pilots in the two airlines. The Wall Street Journal reports that United Airlines tie-
the highest labor cost of all airlines. The information given in these exercises
cicios, does it help to understand why it can be expected that United Airlines will have the costs
higher?
17. The International Air Transport Association conducts surveys among business travelers in the
the quality of international departure airports is rated. The highest rating is 10.
A simple random sample of 50 business travelers was selected, and each was asked about their
Rating for Miami International Airport. The ratings given by these 50 travelers.
The ones shown below.

file CD
6
7
4
8
6
7
8
5
7
9
7
5
6
8
3
4
3
3
8
8
10
5
4
5
8
4
in 4 4 8 4 5 6 2 5 9 9 8 4 8
Miami 9 9 5 9 7 8 3 10 8 9 6

Calculate the 95% confidence interval for the population mean of the grades.
Miami airport.

file CD
During the summer of 2000, 30 fast food restaurants were visited among which were

in contraband Wendy's, McDonald's and Burger King (The Cincinnati Enquirer, July 9, 2000).
The time that elapsed between the customer placing their order and the receipt of the same was recorded.
Fast Food I. The times in the 30 visited restaurants were as follows:

0.9 1.0 1.2 2.2 1.9 3.6 2.8 5.2 1.8 2.1
6.8 1.3 3.0 4.5 2.8 2.3 2.7 5.7 4.8 3.5
2.6 3.3 5.0 4.0 7.2 9.1 2.8 3.6 7.3 9.0

a. Provide a point estimate of the population mean.

b. What is the margin of error with 95% confidence?
c. What is the 95% confidence interval estimate for the population mean?
d. Analyze the bias that may be found in this population. What would you suggest for the repetition?
of this study?
19. A study by the National Retail Foundation found that families were willing to
spend an average of $649 during the December holidays (The Wall Street Journal, December 2)
December 2002). Suppose that 600 families participated in the study and that the standard deviation is -
Sample value was $175.
a. With 95% confidence, what is the margin of error?
b. What is the 95% confidence interval for estimating the population mean?
c. Last year, the average expenditure per family was $632. Analyze the variation.
change in spending during the December holidays in this period of a year.
316 Chapter 8 Interval Estimation

file CD
20. Do commercials constantly interrupt your favorite television show? CNBC pre-
it provided statistical data on the average number of program minutes in half an hour of
in transmission (CNBC, February 23, 2006). The following data (in minutes) are representative-
Program you of your findings.
21.06 22.24 20.62
21.66 21.23 23.86
23.82 8:30 PM 21.52
21.52 21.91 23.14
20.02 22.20 21.20
22.37 22.19 22.34
23.36 23.44
Assume that the population is approximately normal. Provide a point estimate and an interval.
95% confidence interval for the average number of program minutes in half an hour of broadcasting.
21. The consumption of women of legal drinking age has increased in the United Kingdom.
file CD of, the United States and Europe (The Wall Street Journal, February 15, 2006). Data (of consu-
in my annual in liters) reported by The Wall Street Journal found in a sample of 20 women
Alcohol young Europeans are:
266 82 199 174 97
170 222 115 130 169
164 102 113 171 0
93 0 93 110 130
If the population is more or less symmetrical, provide a 95% confidence interval for consumption.
annual average of alcoholic beverages among young European women.
22. The first weeks of 2004 were good for the stock market. In a sample of
25 open funds reported the following gains obtained since the beginning of the year to
January 24, 2004 (Barron’s, January 19, 2004).

file CD
7.0
2.5
3.2
2.5
1.4
1.9
5.4
5.4
8.5
1.6
in 1.0 2.1 8.5 4.3 6.2
Open End Funds 1.5 1.2 2.7 3.8 2.0
1.2 2.6 4.0 2.6 0.6
a. What is the point estimate of the population mean of the earnings in open-end funds?
from the beginning of the year until that date?
b. Since the population has a normal distribution, calculate a confidence interval of
95% for the population mean of earnings in open-end funds since the beginning of the year
until that date.

8.3Determination of sample size

In the practical recommendations of the two previous sections, the role of size was discussed.
from the sample to obtain a good approximation of the confidence intervals in cases where
that the population does not have a normal distribution, attention is now focused on another aspect re-
related to the sample size. It describes how to choose a sufficient sample size.
The procedure that is greatly fearing to obtain a desired margin of error. To explain this, it returns to the case of
present in this section is section 8.1 in which there was a known. With expression (8.1), the confidence interval
used to determine the the action is given by
sample size that is
needs to have a σ
determined margin of xbar Ϯ zα/2
͙ n
error that has been established
before taking the sample. The quantityα/2(σ͞ ͙ n) is the margin of error. So, as seen, zα/2 , the deviation is-
population standard deviation, and the sample size combine to determine the margin of
318 Chapter 8 Interval Estimation

The equation (8.3) I found that the population standard deviation of the daily rental cost was $9.65. Using $9.65
provide the size of as the planned value of σ, we have
minimum necessary sample
to obtain the margin of (zα/2)2σ 2 (1.96)2(9.65)2
desired error. If the size nϭ ϭ ϭ 89.43
sample calculated no
E2 22
it is an integer In this way, the sample size needed to obtain a margin of error of $2 must
round to the next
to be at least 89.43 rentals of medium cars. In cases like this, where the va-
integer, with which
there will be a margin of If denno is an integer, it is rounded up to the next whole number; thus the sample size
error slightly less than The recommended rate is 90 medium car rentals.
required.

Exercises

Methods
23. How large should a sample be selected to have a 95% confidence interval?
with a margin of error of 10? Assume that the population standard deviation is 40.
24. In a dataset, it is estimated that the range is 36.
Carexam a. What is the planned value for the population standard deviation?
b. What size should the sample be so that the margin of error in an interval of
Is 95% confidence 3?
c. What size should the sample be for the margin of error in an interval of
Is the 95% confidence level 2?

Applications
25. Refer to the example of Scheer Industries in section 8.2. As a planned value for the de-
Carexam Standard population deviation uses 6.84 days.
a. At a 95% confidence level, what size should the sample be to have a margin of
1.5 days error?
b. If a 90% confidence interval is desired, what size should the sample be to te-
to have a margin of error of 2 days?
26. The average cost of unleaded gasoline in Greater Cincinnati is $2.41 (The Cincinnati Enquirer,
February 3, 2006). In a time of price changes, a newspaper samples the gasoline-
Prepare and present a report on gasoline prices. Assume that in the prices of the gallon of
The standard deviation of unleaded gasoline is $0.15; give the sample size that should be used.
newspaper to have 95% confidence with each of the following margins of error.
a. A margin of error of $0.07
b. A margin of error of $0.05
c. A margin of error of $0.03
27. The starting annual salaries of students who have just completed a degree in management
It is expected to be between $30,000 and $45,000. Suppose you want to provide a range of
95% confidence interval to estimate the population mean of starting salaries. What is the value?
planned from the population standard deviation? How large should the sample be if you want it to ...
the margin of error is
a. $500?
$200?
$100?
d. Would you recommend trying to have $100 as a margin of error?
28. Smith Travel Research provides information on nightly room rates.
from the hotel in the United States (USAToday, July 8, 2002). Use $2 as the margin of error.
seado and $22.50 as the planned value for the population standard deviation, and find the ta-
sample hands that are requested in sections a, b and c.
a. For a 90% confidence interval, estimate the average price of hotel rooms.
b. For a 95% confidence interval, estimate the average price of hotel rooms.
 8.4 Population proportion 319

c. For a 99% confidence interval, estimate the average price of hotel rooms.
d. When there is a fixed margin of error, what happens to the sample size as it increases?
Does the level of confidence increase? Would you recommend Smith Travel Research to use 99%?
How about the level of confidence? Discuss.

29. The 2003 Information Please Almanac presents the times it takes to commute to work.
required in the 15 largest cities in the United States. Assume that you use a random sample
simple preliminary history of the inhabitants of San Francisco and as a planned value for the deviation-
the standard population completion time is 6.25 minutes.
a. If you want to estimate the population mean of the time they need in San Francisco to trans-
arrive at work, with a margin of error of 2 minutes, what should be the size of the
sample? Assume that the confidence level is 95%.
b. If you want to estimate the population mean of the time needed in San Francisco to
commute to work, with a margin of error of 1 minute, what should be the size of the
sample? Assume that the confidence level is 95%.
In the first quarter of 2003, the price/earnings ratio P/E in the stock market
New York went from 5 to 60 (The Wall Street Journal, March 7, 2003). If you want to estimate the
population average of this P/E ratio in all actions of the New York Stock Exchange, how
What actions must be taken in the sample if one wants the margin of error to be 3? Use 95%.
trustworthy.

8.4Population proportion
In the introduction to this chapter, it was stated that to obtain an interval estimate of the pro-
population portion, the general formula was

p̄mngariof
Ϯ error

In the calculation of this interval estimate, the sampling distribution is important.

In chapter 7, it was said that the sample distribution of p-hat is approximated by a distribution.
normal action always quenpՆ 5 yn(1Ϫ p)Ն 5. Figure 8.9 shows an approximation

FIGURE 8.9NORMAL APPROXIMATION TO THE SAMPLING DISTRIBUTION OF p¯

Sampling distribution p(1 – p)

dep σ p =
n

α /2 α/2

p
p
zα /2σ p zα /2σ p
322 Chapter 8 Interval Estimation

TABLE 8.5SOME VALUES OF p*(1Ϫ p*)

p* p*(1؊ p*)
0.10 (0.10)(0.90)ϭ 0.09
0.30 (0.30)(0.70)ϭ 0.21
0.40 (0.40)(0.60)ϭ 0.24
0.50 (0.50)(0.50)ϭ 0.25 Greater value dep*(1Ϫ p*)
0.60 (0.60)(0.40)ϭ 0.24
0.70 (0.70)(0.30) ϭ 0.21
0.90 (0.90)(0.10) ϭ 0.09

Thus, the sample size must be at least 1514.5 golfers to have the margin of error.
required. Rounding up to the next whole number, it is necessary to have 1515 golfers to
obtain the desired margin of error.
The fourth suggested alternative to select a planned value is to chooseϭ 0.50.
When no other information is available, this value is usually used. To understand the ra-
zone, note that the numerator of equation (8.7) indicates that the sample size is pro-
proportional to the amount p*(1Ϫ p*). If the value dep*(1Ϫ p*) is large, the size of the sample
it will be great. Table 8.5 shows some of the values that p*(1 can haveϪ p*). The
higher value dep*(1Ϫ p*) is presented when p*ϭ 0.50. In this way, in case of doubt, approach-
of the appropriate planned value, it is known thatϭ 0.50 will give the largest sample size that you
can recommend. Indeed, recommending the largest possible sample size goes to
Sure. If it turns out that the sample proportion is different from the planned value, the margin of error
it will be lower than desired. So when using it*ϭ 0.50, it is guaranteed that the size of the
the sample will be sufficient to obtain the desired margin of error.
In the example of the study of the female golfers, if it is used as the planned value.ϭ 0.50, the theme-
the sample obtained is
2
α/2) p*(1p*) (1.96)2(0.50)(1 0.50)
n 1536.6
2
E 0.0252

That is to say, a slightly larger sample, 1537 golfers.

NOTES AND COMMENTS

The desired margin of error for calculating a pro- provides a sample size that is
The population portion is almost always 0.10 or less. sufficient to meet the requirements of
In national public opinion surveys npՆ 5 yn(1Ϫ p)Ն 5 to use a distribution
national in the United States, conducted by Gallup and normal as an approximation of the distribution
Harris, a margin of error of 0.03 or 0.04 is co- sample mean.
mún. With such margins of error, the equation

Exercises

Methods
In a simple random sample of 400 individuals, 100 of the responses were Yes.
Carexam a. Provide the point estimate of the population proportion of individuals whose
b. Give the estimation of the standard error of the proportion σp̄ .
response will be Yes.

c. Calculate the 95% confidence interval for the population proportion.

 8.4 Population proportion 323

32. In a random sample of 800 elements, a sample proportion is obtained,ϭ

p̄ 0.70.
a. Provide a 90% confidence interval for the population proportion.
b. Find a 95% confidence interval for the population proportion.
33. In a study, the planned value for the population proportion esp*ϭ 0.35. What size is it?
should take the sample to give a 95% confidence interval with a margin of error of 0.05?
34. For 95% confidence, what sample size should be taken to achieve a margin of error?
of 0.03 in the estimation of a population proportion? Assume that no data is available.
interiors to obtain a planned value dep*.

Applications
35. A study was conducted with 611 office workers to investigate their attention to the phone, the study records
Carexam the frequency with which they answered the phone and the frequency with which they let the call go unanswered
went to voicemail (USAToday, April 21, 2002). Of these office workers, 281 indicated con-
always try the calls and do not use the voicemail.
a. Give the point estimate of the population proportion of office workers who always respond
the telephone.
b. At 90% confidence, what is the margin of error?
c. Give the 90% confidence interval for the proportion of the population of office workers that
They always answer the phone.
36. According to statistics published by CNBC, the number of vehicles that are not affordable
guarantees are surprising (CNBC, February 23, 2006). The sample results from CNBC
It indicates that 46 out of 200 vehicles were not insured.
a. What is the point estimate of the proportion of uninsured vehicles?
b. Give a 95% confidence interval for the population proportion.
37. Towers Perrin, a human resources company, conducted a study with 1100 employees of em-
medium and large companies to determine how dissatisfied they were with their jobs
Wall Street Journal, January 29, 2003). The JobSatisfaction file shows representative data.
Affirmative. A Yes as a response indicates that the employee is very displeased with their current job.
a. Give the point estimate of the population proportion of employees who dislike it.
file CD
much about his current job.
in b. At 95% confidence, what is the margin of error?
c. What is the 95% confidence interval for the proportion of the employee population?
Job Satisfaction
Who dislikes their current job very much?
d. Towers Perrin estimates that it costs employers one third of an annual salary per hour.
find a successor and up to 1.5 times the annual salary find a successor for an employee
who receives a high compensation. What is the message of this research for the
employers?
38. According to Thomson Financial, as of January 25, 2006, most of the companies that in-
they were expected to have profits that surpassed estimates (BusinessWeek, February 6th
In a sample of 162 companies, 104 exceeded the estimates, 29 matched, and 29 did not.
they fell short.
a. What is the point estimate of the proportion of companies that fell short?
b. Determine the margin of error and give a 95% confidence interval for the proportion that
exceeded the estimates.
c. What size should the sample be if the margin of error is 0.05?
In 2003, the percentage of people without health insurance (in the United States) was 15.6%.
Carexam (Statistical Abstract of the United States, 2006). A committee of Congress is asked to carry out a
study to obtain updated information.
a. What sample size do you recommend to the committee, if the objective is to estimate
the estimation of the current proportion of individuals who do not have health insurance the margin of error
sea 0.03? Use 95% confidence.
b. Repeat item a using 99% confidence.
324 Chapter 8 Interval Estimation

40. In professional baseball, the record for home runs was 61 hit in a season and
belonged for 37 years to Roger Maris of the New York Yankees. However, since 1998
In 2001, three players —Mark McGwire, Sammy Sosa, and Barry Bonds— broke this record.
Bonds is the one who holds the current record of 73 home runs in a single season. Due to the
breaking this record after such a long time, and also the breaking of other records,
It is suspected that baseball players use illegal drugs known as this-
steroids for muscle formation. In a survey conducted by USAToday/CNN/Gallup
found that 86% of baseball fans believed that professional players were
should undergo a test to detect steroids (USAToday, July 18, 2002). If in a
study selects 650 baseball fans, calculate the margin of error and the interval of...
95% confidence for the proportion of fans who believe that the players of
Professional baseball players should undergo testing to detect steroids.
41. The youth of the United States uses the Internet intensely; 87% of young people between the ages of 12 and 17
years are Internet users (The Cincinnati Enquirer, February 7, 2006). In a sample
of Internet users of this age, 9% voted for MySpace as the most popular website
Suppose that 1,400 young people participated in this study. What are the margins of error and
the interval estimation of the population proportion of those who consider this site to be
the most popular? Use 95% confidence.
42. A survey conducted by USAToday/CNN/Gallup during the presidential campaign took into account
Ninety-one out of a sample of 491 potential voters (USAToday, June 9, 2000). The goal of this...
This survey was to estimate the proportion of potential voters in favor of each candidate. Assume
that the planned value esp*ϭ 0.50 with a confidence level of 95%.
a. If p*ϭ 0.50. What was the planned margin of error in the June survey?
As the November election approaches, a better accuracy and a smaller margin are being sought.
error. Assume that the error margins requested are those shown in the table
next. Calculate the sample size recommended for each study.

Study Margin of error

September 0.04
October 0.03
Beginning of November 0.02
One day before the election 0.01

43. Phoenix Wealth Management/Harris conducted a study with 1500 individuals whose wealth was
of a million or more dollars, obtained various statistics about rich people (BusinessWeek,
September 22, 2003). The previous three years had been bad for the stock market.
yes, which motivated some of the questions asked.
a. This study found that 53% of respondents lost 25% or more of the value of their
portfolio in the last three years. Provide a 95% confidence interval for the proportion
of wealthy people who lost 25% or more of the value of their portfolio in the last three years.
b. The study indicated that 31% of respondents feel that they will need to save more for their retirement to
make up for what was lost. Give a 95% confidence interval for the population proportion.
c. Five percent of the respondents made a donation of $25,000 or more for works of
charity last year. Provide a 95% confidence interval for the proportion of those who hi-
They made a donation of $25,000 or more for charity work.
d. Compare the margins of error of the interval estimates for items a, b, and c.
What is the relationship between the margin of error and p̄? If you use the same sample to obtain several
proportions, which of the proportions should be used to estimate the planned value*? Why
What do you consider is usually used in these cases?ϭ 0.50?

Summary
In this chapter, the methods for obtaining interval estimates for the mean were presented.
blational and of the population proportion. A point estimator may or may not provide a good
in the estimation of a population parameter. An estimation interval provides a measure of
326 Chapter 8 Interval Estimation

Key formulas

Interval estimation of a population mean: known σ

σ
x̄ Ϯ zα/2 (8.1)
͙ n
Interval estimation of a population mean: unknown σ

xbar Ϯ tα/2 (8.2)

͙ n

Sample size for estimating the population mean

(zα/2)2σ 2
nϭ (8.3)
E2
Interval estimation for a population proportion

ͱp¯(1Ϫ p¯)
p̄ Ϯ zalpha over 2
n
(8.6)

Sample size for estimating the population proportion

2
α/2) p*(1Ϫ p*)
nϭ (8.7)
E2

Complementary exercises

44. In a study conducted with 54 discount brokers, it was found that the average of
the prices charged for a transaction of 100 shares at $50 per share was $33.77 (AAII Jour-
January, February 2006). This study is conducted annually. According to historical data
available, consider that the known population standard deviation is $15.
a. According to the sample data, what is the margin of error in a 95% confidence interval?
b. Give a 95% confidence interval for the mean of the prices charged for a transaction.
purchase of 100 shares at $50 per share.
In a study conducted by the American Automobile Association, it was found that a family of
four members on vacation spend an average of $215.60 per day. Suppose that in a sample
of 64 families on vacation at Niagara Falls, the sample mean found is
do $252.45 per day and the sample standard deviation $74.50.
a. Provide an estimate, using a 95% confidence interval for the mean of the amount.
What does a family of four spend per day on vacation at Niagara Falls?
b. According to the confidence interval from part a, does it seem that the population mean of
the amount spent per day by families visiting Niagara Falls is different
What is the media reported by the American Automobile Association? Explain.
46. The movie Harry Potter and the Philosopher's Stone broke the box office opening record of The
lost world: Jurassic Park (The Wall Street Journal, November 19, 2001). In a
A sample of 100 cinemas found that the average gross revenue in the three days of the first
The weekend was $25,467 for the cinema. The standard deviation was $4,980.
a. What is the margin of error in this study? Use 95% confidence.
b. What is the 95% confidence interval estimate of the population mean of the re-
fundraising through cinema?
The Lost World grossed $72.1 million in the three days of its opening weekend. Harry
Potter and the Philosopher's Stone was released in 3672 theaters. Provide an estimate of the total obtained.
Harry Potter and the Philosopher's Stone in the three days of the first weekend.
d. An Associated Press article stated that Harry Potter "shattered" the record of
box office in the debut of The Lost World. Do its results match this?
 Complementary exercises 327

47. Many stock market observers claim that when the P/E ratio in the
if actions is above 20, the market is overvalued. The P/E ratio is the price of a
action divided by the earnings of the last 12 months. Assume you want to know if ac-
Currently, the market is overvalued and what proportion of companies pay dividends.
Here is a list of 30 companies that are listed on the New York Stock Exchange (NYSE).
(Barron’s, January 19, 2004).

Company Dividends Proportion P/E Company Dividends Proportion P/E

Albertsons Yes 14 NY Times A Yes 25
BRE Prop Yes 18 Omnicare Yes 25
CityNtl Yes 16 PallCp Yes 23
DelMonte No 21 Public Service Enterprise
Yes 11

file CD
EnrgzHldg
Ford Motor
No
Yes
20
22
SensientTch
SmtProp
Yes
Yes
11
12
in Gildan A No
Hudson United BCP Yes
12
13
TJX Companies
Thomson
Yes
Yes
21
30
NYSE Stocks IBM Yes 22 USB Hldg Yes 12
JeffPilot Yes 16 US Restr Yes 26
KingswayFin No 6 VarianMed No 41
Libbey Yes 13 Visx No 72
Masonite International
No 15 Waste Management No 23
Motorola Yes 68 Wiley A Yes 21
Ntl City Yes 10 Yum Brands No 18

a. Provide a point estimate for the population proportion P/E in the listed stocks.
on the New York Stock Exchange. Provide a 95% confidence interval.
b. According to your answer to item a, do you consider that the market is overvalued?
c. Provide a point estimate of the proportion of companies on the NYSE that pay dividends.
Is the sample size large enough to justify the use of the distribution?
Is it normal to calculate a confidence interval for this proportion? Why or why not?
Why not?

file CD
USAirways conducted various studies indicating significant savings if travelers frequently...

in The Dividend Miles program will conduct the redemption of miles online and will schedule flights.
livestock (US Airways Attaché, February 2003). In a study, data was collected on the time-
Flights What is required to exchange miles and book a flight by phone?
needed. In the Flights dataset, there is a sample of times in minutes required
to program by phone each of 150 flights won. Use Minitab or Excel to answer
the following questions.
a. What is the sample mean of the number of minutes required to program by telephone?
Is the flight won?
b. Give the 95% confidence interval for the population mean of the time required for
to book flights by phone.
c. Suppose that a telephone ticket agent works 7.5 hours. How many flights won?
Is it expected that the agent will attend to this in a day?
d. Say how this information supports US Airways' plan to use an online system for
reduce costs.

file CD
49. In a study, 200 executives from a sample were asked to provide data on the amount

in minutes per day that office workers lose trying to locate poorly stored items, poorly ar-
misclassified or poorly classified. The data from this research is found in the dataset.
ActTemps to ActTemps.
a. Use ActTemps to provide a point estimate of the minutes per day lost by the office.
cinistas in locating things poorly stored, poorly archived, or poorly classified.
b. What is the sample standard deviation?
c. Give a 95% confidence interval for the number of minutes lost per day.
50. Gasoline performance tests are conducted with a certain model of car. If they
wants to provide a 98% confidence interval with a margin of error of 1 mile per gallon, how-
328 Chapter 8 Interval Estimation

Which cars should be used? Assume that previous tests have shown that the deviation
the performance standard is 2.6 miles per gallon.
A medical center wants to estimate the average time needed to schedule an appointment.
a patient. What size should the sample be if we want the margin of error to be two
minutes and what should be the confidence level at 95%? What sample size should be taken if
Do you want the confidence level to be 99%? For the population standard deviation use as goes -
planned, 8 minutes.
52.BusinessWeek presents data on the annual salary plus bonuses of chairpersons of boards.
administration. In a preliminary sample, the standard deviation is $675; the data is given in my-
dollars. How many chairmen of boards of directors should the sample include?
If you want to estimate the annual salary plus bonuses with a margin of error of $100,000? (Note: The
desired margin of error will beϭ 100 if the data is provided in thousands of dollars.) Use 95% of
trust.
53. The National Center for Education Statistics reports that 47% of university students
he works to pay for his studies and his livelihood. Suppose a sample of 450 students was employed.
university students in that study.
a. Give a 95% confidence interval for the population proportion of students who bring
they work to support themselves and pay for their studies.
b. Give a 99% confidence interval for the population proportion of students who bring
They work to support themselves and pay for their studies.
c. What happens to the margin of error when the confidence level increases from 95% to 99%?
In a study by USAToday/CNN/Gallup, conducted with 369 working parents, it was found that
200 consider that they spend very little time with their children due to work.
a. Provide a point estimate of the population proportion of parents who work and think
they spend very little time with their children due to work.
b. What is the margin of error for 95% confidence?
c. What is the 95% confidence interval for the population proportion of parents who bring
Do they feel that they spend very little time with their children due to work?
55. Which would be harder for you to get rid of: your television or your computer? In a study re-
A survey conducted with 1677 internet users in the United States found that 74% of them
youth elite (average age of 22 years) would find it harder to get rid of their computer (PC)
Magazine, February 3, 2004). Only 48% would find it more difficult to get rid of their television.
a. Give a 95% confidence interval for the proportion of young people for whom it would be more
hard to get rid of your computer.
b. Find a 99% confidence interval for the proportion of young people to whom it applies
it would be harder to get rid of your television.
c. In which of the items a or b is the margin of error greater? Explain why.
56. The Cincinnati/Northern Kentucky International Airport ranked second in 2005 in terms of...
reality in the arrival of flights between the busiest airports in the country (The Cincinna-
the Enquirer, February 3, 2003). Assume that this is based on a sample of 550 flights from the
which 455 arrived on time.
a. Develop a point estimate of the on-time arrival rate (proportion of flights that
They arrive on time at the airport.
b. Build a 95% confidence interval for the proportion of on-time arrivals in the
population of all flights at the airport in 2005.
57. The 2003 Statistical Abstract of the United States states the percentage of people aged 18 and older.
who smokes. Assume that in a new study to collect data on smokers and non-smokers-
it is used as a preliminary estimate of the proportion that smokes, 0.30.
a. What size should the sample be to estimate the proportion of smokers with a
margin of error of 0.02? Use 95% confidence.
b. Suppose the study uses your recommendation for the sample size from paragraph a and
find 520 smokers. What is the point estimate of the proportion of smokers in
the population?
c. What is the 99% confidence interval for the proportion of smokers in the population?
 Problem Case 1 The Young Professional Magazine 329

A credit card company of a well-known bank wants to estimate the proportion of cardholders
items that at the end of the month have a balance different from zero which causes charges. Suppose that the
The desired margin of error is 0.03 with 98% confidence.
a. What size should the sample be if it is believed that 70% of the cardholders of the
Do they have a balance different from zero at the end of the month?
b. What size should the sample be if no planned value can be given for
the proportion?
59. In a study, 200 people were asked to indicate their main source of news information.
110 indicated that their main source of news was television newscasts.
a. Give a 95% confidence interval for the population proportion of people who have
as the main source of news to television.
b. What will be the necessary sample size to estimate the population proportion with a
margin of error of 0.05 and 95% confidence?
60. Although for business travelers, schedules and costs are important factors when choosing
An airline, in a study conducted by USAToday, it was found that for travelers of ne-
In business, the most important factor is that the airline has a frequent flyer program. In a
show denϭ 1993 passengers who participated in the study, 618 indicated as the most significant factor
important a frequent passenger program.
a. What is the point estimate of the population proportion of business travelers that
they consider the frequent flyer program as the most important factor when choosing an airline.
airway?
b. Give a 95% confidence interval to estimate the population proportion.
c. What size should the sample be so that the margin of error is 0.01 with 95% of
trust? Would you recommend that USAToday try to have this accuracy? Why yes or why not?
what not?

Case problem 1The Young Professional magazine

The magazine Young Professional was created for an audience made up of people who are
During the first 10 years of his professional career in business. In his first two years of
publication, the magazine has been quite successful. Now the editor is trying to increase the base
advertising for your magazine. Potential advertisers continuously ask about the data
demographics and interests of the subscribers of Young Professional. To gather this information
The magazine conducted a study to develop the profile of its subscribers. The results of it-
The study will be used to help select items of interest and to provide to the
advertisers a profile of the subscribers. As a new employee of the company, you are asked to us-
I need your help to analyze the results of the study.
The following are some of the questions from the study.

1.Age___
file CD 2.Sex____
in Are you considering purchasing any real estate in the next two years? Yes__ No__
Professional 4. What is the approximate value of financial investments, excluding your house, that are
property of yours or another member of your household?
5. How many transactions of stocks/bonds/mutual funds did you make last year?
Do you have broadband access to the Internet at home? Yes ___ No ___
7. Please indicate what your household income was last year.
8. Do you have children? Yes__ No___

The Professional file contains the answers to these questions. Table 8.6 shows
the part of this file corresponding to the responses of the first five interviewees. The
The complete file can be found on the compact disc that is distributed with the book.