CHAPTER

16 STATISTICS
Every four years, each major political party in the
United States holds a convention to select the party’s
nominee for President of the United States. Before
CHAPTER these conventions are held, each candidate assembles
TABLE OF CONTENTS a staff whose job is to plan a successful campaign.This
16-1 Collecting Data plan relies heavily on statistics: on the collection and
16-2 Organizing Data
organization of data, on the results of opinion polls,
and on information about the factors that influence
16-3 The Histogram
the way people vote. At the same time, newspaper
16-4 The Mean, the Median, and
the Mode
reporters and television commentators assemble
other data to keep the public informed on the prog-
16-5 Measures of Central Tendency
and Grouped Data ress of the candidates.
16-6 Quartiles, Percentiles, and
Election campaigns are just one example of the use
Cumulative Frequency of statistics to organize data in a way that enables us
16-7 Bivariate Statistics to use available information to evaluate the current sit-
Chapter Summary
uation and to plan for the future.
Vocabulary
Review Exercises
Cumulative Review

660
 Collecting Data 661

16-1 COLLECTING DATA

In our daily lives, we often deal with problems that involve many related items
of numerical information called data. For example, in the daily newspaper we
can find data dealing with sports, with business, with politics, or with the
weather.
Statistics is the study of numerical data. There are three typical steps in a
statistical study:
STEP 1. The collection of data.
STEP 2. The organization of these data into tables, charts, and graphs.
STEP 3. The drawing of conclusions from an analysis of these data.
When these three steps, which describe and summarize the formation and
use of a set of data, are included in a statistical study, the study is often called
descriptive statistics. You will study these steps in this first course. In some cases,
a fourth step, in which the analyzed data are used to predict trends and future
events, is added.
Data can be either qualitative or quantitative. For example, a restaurant
may ask customers to rate the meal that was served as excellent, very good,
good, fair, or poor. This is a qualitative evaluation. Or the restaurant may wish
to make a record of each customer tip at different times of the day. This is a
quantitative evaluation, which lends itself more readily to further statistical
analysis.
Data can be collected in a number of ways, including the following:
1. A written questionnaire or list of questions that a person can answer by
checking one of several categories or supplying written responses.
Categories to be checked may be either qualitative or quantitative.
Written responses are usually qualitative.
2. An interview, either in person or by telephone, in which answers are given
verbally and responses are recorded by the person asking the questions.
Verbal answers are usually qualitative.
3. A log or a diary, such as a hospital chart or an hourly recording of the out-
door temperature, in which a person records information on a regular
basis. This type of information is usually quantitative.

Note: Not all numerical data are quantitative data. For instance, a researcher
wishes to investigate the eye color of the population of a certain island. The
researcher assigns “blue” to 0, “black” to 1, “brown” to 2, and so on. The
resulting data, although numerical, are qualitative since it represents eye color
and the assignment was arbitrary.
662 Statistics

Sampling
A statistical study may be useful in situations such as the following:
1. A doctor wants to know how effective a new medicine will be in curing a
disease.
2. A quality-control team wants to know the expected life span of flashlight
batteries made by its company.
3. A company advertising on television wants to know the most frequently
watched TV shows so that its ads will be seen by the greatest number of
people.
When a statistical study is conducted, it is not always possible to obtain
information about every person, object, or situation to which the study applies.
Unlike a census, in which every person is counted, some statistical studies use
only a sample, or portion, of the items being investigated.
To find effective medicines, pharmaceutical companies usually conduct tests
in which a sample, or small group, of the patients having the disease under study
receive the medicine. If the manufacturer of flashlight batteries tested the life
span of every battery made, the warehouse would soon be filled with dead bat-
teries. The manufacturer tests only a sample of the batteries to determine their
average life span. An advertiser cannot contact every person owning a TV set to
determine which shows are being watched. Instead, the advertiser studies TV
ratings released by a firm that conducts polls based on a small sample of TV
viewers.
For any statistical study, whether based on a census or a sample, to be use-
ful, data must be collected carefully and correctly. Poorly designed sampling
techniques result in bias, that is, the tendency to favor a selection of certain
members of the population which, in turn, produces unreliable conclusions.

Techniques of Sampling
We must be careful when choosing samples:
1. The sample must be fair or unbiased, to reflect the entire population being
studied. To know what an apple pie tastes like, it is not necessary to eat the
entire pie. Eating a sample, such as a piece of the apple pie, would be a fair
way of knowing how the pie tastes. However, eating only the crust or only
the apples would be an unfair sample that would not tell us what the
entire pie tastes like.
2. The sample must contain a reasonable number of the items being tested or
counted. If a medicine is generally effective, it must work for many people.
The sample tested cannot include only one or two patients. Similarly, the
manufacturer of flashlight batteries cannot make claims based on testing
five or 10 batteries. A better sample might include 100 batteries.
 Collecting Data 663

3. Patterns of sampling or random selection should be employed in a study.

The manufacturer of flashlight batteries might test every 1,000th battery to
come off the assembly line. Or, the batteries to be tested might be selected
at random.
These techniques will help to make the sample, or the small group, repre-
sentative of the entire population of items being studied. From the study of the
small group, reasonable conclusions can be drawn about the entire group.

EXAMPLE 1

To determine which television programs are the most popular in a large city, a
poll is conducted by selecting people at random at a street corner and inter-
viewing them. Outside of which location would the interviewer be most likely to
find an unbiased sample?
(1) a ball park (2) a concert hall (3) a supermarket

Solution People outside a ball park may be going to a game or purchasing tickets for a
game in the future; this sample may be biased in favor of sports programs.
Similarly, those outside a concert hall may favor musical or cultural programs.
The best (that is, the fairest) sample or cross section of people for the three
choices given would probably be found outside a supermarket.

Answer (3)

Experimental Design
So far we have focused on data collection. In an experiment, a researcher
imposes a treatment on one or more groups. The treatment group receives the
treatment, while the control group does not.
For instance, consider an experiment of a new medicine for weight loss.
Only the treatment group is given the medicine, and conditions are kept as sim-
ilar as possible for both groups. In particular, both groups are given the same
diet and exercise. Also, both groups are of large enough size and are chosen such
that they are comprised of representative samples of the general population.
However, it is often not enough to have just a control group and a treatment
group. The researcher must keep in mind that people often tend to respond to
any treatment. This is called the placebo effect. In such cases, subjects would
report that the treatment worked even when it is ineffective. To account for
the placebo effect, researchers use a group that is given a placebo or a dummy
treatment.
Of course, subjects in the experimental and placebo groups should not know
which group they are in (otherwise, psychology will again confound the results).
The practice of not letting people know whether or not they have been given the
real treatment is called blinding, and experiments using blinding are said to be
single-blind experiments. When the variable of interest is hard to measure or
664 Statistics

define, double-blind experiments are needed. For example, consider an experi-

ment measuring the effectiveness of a drug for attention deficit disorder. The
problem is that “attention deficiency” is difficult to define, and so a researcher
with a bias towards a particular conclusion may interpret the results of the
placebo and treatment groups differently. To avoid such problems, the
researchers working directly with the test subjects are not told which group a
subject belongs to.

Interpreting Graphs of Data

Oftentimes embellishments to graphs distort the perception of the data, and so
you must exercise care when interpreting graphs of data.
1. Two- and three-dimensional figures.
As the graph on the right CRIME RATE IN THE U.S.
shows, graphs using two- or
three-dimensional figures can 1990 = 14,475,613
distort small changes in the
POLICE
data. The lengths show the
decrease in crime, but since our
eyes tend to focus on the areas,
the total decrease appears 1995 = 13,862,727
greater than it really is. The rea- POLICE
son is because linear changes
are increased in higher dimen-
sions. For instance, if a length
doubles in value, say from x to
2x, the area of a square with 2000 = 11,876,669
POLICE
sides of length x will increase by
x2 → (2x)2  4x2,
a four-fold increase. Similarly, the volume of a cube with edges of length x
will increase by
x3 → (2x)3  8x3,
an eight-fold increase!
2. Horizontal and vertical scales.
The scales used on the vertical and horizontal axes can exaggerate,
diminish, and/or distort the nature of the change in the data. For instance,
in the graph on the left of the following page, the total change in weight is
less than a pound, which is negligible for an adult human. However, the
scale used apparently amplifies this amount. While on the right, the
unequal horizontal scale makes the population growth appear linear.
 Collecting Data 665

AVERAGE WEIGHT OF SUBJECTS

OVER 6-MONTH PERIOD
200.0
199.9 POPULATION OF ANYTOWN, U.S.
199.8
199.7 500
Weight (lb)

(in thousands)
199.6

Population
400
199.5
199.4 300
199.3
199.2 200
199.1 100
199.0

19

19

19

19

20

20

20
M

80

85

90

99

00

02

06
on

on

on

on

on

on
Year
th

th

th

th

th

th
1

6
Month

EXERCISES
Writing About Mathematics
1. A census attempts to count every person. Explain why a census may be unreliable.
2. A sample of a new soap powder was left at each home in a small town. The occupants were
asked to try the powder and return a questionnaire evaluating the product. To encourage
the return of the questionnaire, the company promised to send a coupon for a free box of
the soap powder to each person who responded. Do you think that the questionnaires that
were returned represent a fair sample of all of the persons who tried the soap? Explain why
or why not.

Developing Skills
In 3–10, determine if each variable is quantitative or qualitative.
3. Political affiliation 4. Opinions of students on a new music album
5. SAT scores 6. Nationality
7. Cholesterol level 8. Class membership (freshman, sophomore, etc.)
9. Height 10. Number of times the word “alligator” is used in an essay.

In 11–18, in each case a sample of students is to be selected and the height of each student is to be
measured to determine the average height of a student in high school. For each sample:
a. Tell whether the sample is biased or unbiased.
b. If the sample is biased, explain how this might affect the outcome of the survey.
11. The basketball team 12. The senior class
13. All 14-year-old students 14. All girls
666 Statistics

15. Every tenth person selected from an alphabetical list of all students
16. Every fifth person selected from an alphabetical list of all boys
17. The first three students who report to the nurse on Monday
18. The first three students who enter each homeroom on Tuesday

In 19–24, in each case the Student Organization wishes to interview a sample of students to deter-
mine the general interests of the student body. Two questions will be asked: “Do you want more pep
rallies for sports events? Do you want more dances?” For each location, tell whether the Student
Organization would find an unbiased sample at that place. If the sample is biased, explain how this
might influence the result of the survey.
19. The gym, after a game 20. The library
21. The lunchroom 22. The cheerleaders’ meeting
23. The next meeting of the Junior Prom committee
24. A homeroom section chosen at random
25. A statistical study is useful when reliable data are collected. At times, however, people may
exaggerate or lie when answering a question. Of the six questions that follow, find the three
questions that will most probably produce the largest number of unreliable answers.
(1) What is your height? (2) What is your weight?
(3) What is your age? (4) In which state do you live?
(5) What is your income? (6) How many people are in your family?
26. List the three steps necessary to conduct a statistical study.
27. Explain why the graph below is misleading.

SUMMER OLYMPIC GAMES CHAMPIONS

100-METER RACE

1988 1992 1996 2000 2004

Carl Lewis Linford Christie Donovan Bailey Maurice Green Justin Gatlin
(USA) (GBR) (CAN) (USA) (USA)
9.92 sec 9.96 sec 9.84 sec 9.87 sec 9.85 sec
 Organizing Data 667

28. Investigators at the University of Kalamazoo were interested in determining whether or not
women can determine a man’s preference for children based on the way that he looks.
Researchers asked a group of 20 male volunteers whether or not they liked children. The
researchers then showed photographs of the faces of the men to a group of 10 female volun-
teers and asked them to pick out which men they thought liked children. The women cor-
rectly identified over 90% of the men who said they liked children. The researchers
concluded that women could identify a man’s preference for children based on the way that
he looks. Identify potential problems with this experiment.

Hands-On Activity
Collect quantitative data for a statistical study.
1. Decide the topic of the study. What data will you collect?
2. Decide how the data will be collected. What will be the source(s) of that data?
a. Questionnaires
b. Personal interviews
c. Telephone interviews
d. Published materials from sources such as almanacs or newspapers.
3. Collect the data. How many values are necessary to obtain reliable information?
Keep the data that you collect to use as you learn more about statistical studies.

16-2 ORGANIZING DATA

Data are often collected in an unorganized and random manner. For example, a
teacher recorded the number of days each of 25 students in her class was absent
last month. These absences were as follows:
0, 3, 1, 0, 4, 2, 1, 3, 5, 0, 2, 0, 0, 0, 4, 0, 1, 1, 2, 1, 0, 7, 3, 1, 0
How many students were absent fewer than 2 days? What was the number
of days for which the most students were absent? How many students were
absent more than 5 days? To answer questions such as these, we find it helpful
to organize the data.
One method of organizing data is to write it as an ordered list. In order from
least to greatest, the absences become:
0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 1, 1, 1, 1, 1, 2, 2, 2, 3, 3, 3, 4, 4, 5, 7
We can immediately observe certain facts from this ordered list: more students
were absent 0 days than any other number of days, the same number of students
were absent 5 and 7 days. However, for more a quantitative analysis, it is useful
to make a table.
668 Statistics

Preparing a Table
In the left column of the accompanying Absences Tally
table, we list the data values (in this case
the number of absences) in order. We start 7
with the largest number, 7, at the top and
go down to the smallest number, 0. 6
For each occurrence of a data value, we
place a tally mark, |, in the row for that 5
number. For example, the first data value in
4
the teacher’s list is 0, so we place a tally in
the 0 row; the second value is 3, so we place 3
a tally in the 3 row. We follow this proce-
dure until a tally for each data value is 2
recorded in the proper row. To simplify
counting, we write every fifth tally as a 1
diagonal mark passing through the first
four tallies: . 0
Once the data have been orga-
nized, we can count the number of tally
marks in each row and add a column Absences Tally Frequency
for the frequency, that is, the number of
times that a value occurs in the set of 7 1
data. When there are no tally marks in
6 0
a row, as for the row showing 6
absences, the frequency is 0.The sum of 5 1
all of the frequencies is called the total
frequency. In this case, the total fre- 4 2
quency is 25. (It is always wise to check
the total frequency to be sure that no 3 3
data value was overlooked or dupli-
cated in tallying.) From the table, called 2 3
a frequency distribution table, it is now
1 6
easy to see that 15 students were absent
fewer than 2 days, that more students 0 9
were absent 0 days (9) than any other
number of days, and that 1 student was Total frequency 25
absent more than 5 days.

Grouped Data
A teacher marked a set of 32 test papers. The grades or scores earned by the stu-
dents were as follows:
90, 85, 74, 86, 65, 62, 100, 95, 77, 82, 50, 83, 77, 93, 73, 72,
98, 66, 45, 100, 50, 89, 78, 70, 75, 95, 80, 78, 83, 81, 72, 75
 Organizing Data 669

Because of the large number of different scores, it is convenient to orga-

nize these data into groups or intervals, which must be equal in size. Here
we will use six intervals: 41–50, 51–60, 61–70, 71–80, 81–90, 91–100. Each inter-
val has a length of 10, found by subtracting the starting point of an interval
from the starting point of the next
higher interval.
For each test score, we now Interval Tally Frequency
place a tally mark in the row for the
91–100 6
interval that includes that score. For
example, the first two scores in the 81–90 8
list above are 90 and 85, so we place
two tally marks in the interval 71–80 11
81–90. The next score is 74, so we
place a tally mark in the interval 61–70 4
71–80. When all of the scores have
been tallied, we write the frequency 51–60 0
for each interval.
This table, containing a set of 41–50 3
intervals and the corresponding fre-
quency for each interval, is an exam-
ple of grouped data.
When unorganized data are grouped into intervals, we must follow certain
rules in setting up the intervals:

1. The intervals must cover the complete range of values. The range is the
difference between the highest and lowest values.
2. The intervals must be equal in size.
3. The number of intervals should be between 5 and 15. The use of too many
or too few intervals does not make for effective grouping of data. We usu-
ally use a large number of intervals, for example, 15, only when we have a
very large set of data, such as hundreds of test scores.
4. Every data value to be tallied must fall into one and only one interval.
Thus, the intervals should not overlap. When an interval ends with a
counting number, the following interval begins with the next counting
number.
5. The intervals must be listed in order, either highest to lowest or lowest to
highest.
670 Statistics

These rules tell us that there are

many ways to set up tables, all of Interval Tally Frequency
them correct, for the same set of
93–100 6
data. For example, here is another
correct way to group the 32 unorga- 85–92 4
nized test scores given at the begin-
ning of this section. Note that the 77–84 9
length of the interval here is 8.
69–76 7

61–68 3

53–60 0

45–52 3

Constructing a Stem-and-Leaf Diagram

Another method of displaying data is called a stem-and-leaf diagram. The stem-
and-leaf diagram groups the data without losing the individual data values.
A group of 30 students were asked to record the length of time, in minutes,
spent on math homework yesterday. They reported the following data:
38, 15, 22, 20, 25, 44, 5, 40, 38, 22, 20, 35, 20, 0, 36,
27, 37, 26, 33, 25, 17, 45, 22, 30, 18, 48, 12, 10, 24, 27
To construct a stem-and-leaf diagram for the lengths of time given, we begin
by choosing part of the data values to be the stem. Since every score is a one- or
two-digit number, we will choose the tens digit as a convenient stem. For the
one-digit numbers, 0 and 5, the stem is 0; for the other data values, the stem is
1, 2, 3, or 4. Then the units digit will be the leaf. We construct the diagram as
follows:
STEP 1. List the stems, starting with 4, under one another Stem Leaf
to the left of a vertical line beneath a crossbar. 4
3
2
1
0

STEP 2. Enter each score by writing its leaf (the units Stem Leaf
digit) to the right of the vertical line, following 4
the appropriate stem (its tens value). For exam- 3 8
ple, enter 38 by writing 8 to the right of the verti- 2
cal line, after stem 3. 1
0
 Organizing Data 671

STEP 3. Add the other scores to the dia- Stem Leaf

gram until all are entered. 4 4058
3 8856730
2 205200765247
1 57820
0 50

STEP 4. Arrange the leaves in order after Stem Leaf

each stem. 4 0458
3 0356788
2 000222455677
1 02578
STEP 5. Add a key to demonstrate the
0 05
meaning of each value in the
diagram. Key: 3  0  30

EXAMPLE 1

The following data consist of the weights, in kilograms, of a group of 30 students:

70, 43, 48, 72, 53, 81, 76, 54, 58, 64, 51, 53, 75, 62, 84,
67, 72, 80, 88, 65, 60, 43, 53, 42, 57, 61, 55, 75, 82, 71
a. Organize the data in a table. Use five intervals starting with 40–49.
b. Based on the grouped data, which interval contains the greatest number of
students?
c. How many students weigh less than 70 kilograms?

Solution a.
Frequency
Interval Tally (number)

80–89 5

70–79 7

60–69 6

50–59 8

40–49 4

b. The interval 50–59 contains the greatest number of students, 8. Answer

c. The three lowest intervals, namely 40–49, 50–59, and 60–69, show weights
less than 70 kilograms. Add the frequencies in these three intervals:
4  8  6  18 Answer
672 Statistics

EXAMPLE 2

Draw a stem-and-leaf diagram for the data in Example 1.

Solution Let the tens digit be the stem and the units digit the leaf.
(1) Enter the data values in the (2) Arrange the leaves in numerical
given order: order after each stem:

Stem Leaf Stem Leaf

8 14082 8 01248
7 0265251 7 0122556
6 427501 6 012457
5 34813375 5 13334578
4 3832 4 2338
(3) Add a key indicating unit of measure: Key: 5  1  51 kg

EXERCISES
Writing About Mathematics
1. Of the examples given above, which gives more information about the data: the table or the
stem-and-leaf diagram? Explain your answer.
2. A set of data ranges from 2 to 654. What stem can be used for this set of data when drawing
a stem-and-leaf diagram? What leaves would be used with this stem? Explain your choices.

Developing Skills
3. a. Copy and complete the table to group the data, which represent the heights, in centime-
ters, of 36 students:
162, 173, 178, 181, 155, 162, 168, 147, 180,
Interval Tally Frequency
171, 168, 183, 157, 158, 180, 164, 160, 171,
183, 174, 166, 175, 169, 180, 149, 170, 150, 180–189
158, 162, 175, 171, 163, 158, 163, 164, 177
170–179
b. Use the grouped data to answer the following questions:
160–169
(1) How many students are less than 160 centimeters
in height? 150–159
(2) How many students are 160 centimeters or more in 140–149
height?
(3) Which interval contains the greatest number of students?
(4) Which interval contains the least number of students?
 Organizing Data 673

c. Display the data in a stem-and-leaf diagram. Use the first two digits of the numbers as
the stems.
d. What is the range of the data?
e. How many students are taller than 175 centimeters?
4. a. Copy and complete the table to group the data, which gives the lifespan, in hours, of 50
flashlight batteries:
73, 81, 92, 80, 108, 76, 84, 102, 58, 72, Interval Tally Frequency
82, 100, 70, 72, 95, 105, 75, 84, 101, 62,
63, 104, 97, 85, 106, 72, 57, 85, 82, 90, 50–59
54, 75, 80, 52, 87, 91, 85, 103, 78, 79, 60–69
91, 70, 88, 73, 67, 101, 96, 84, 53, 86
70–79
b. Use the grouped data to answer the following questions:
80–89
(1) How many flashlight batteries lasted for 80 or
more hours? 90–99
(2) How many flashlight batteries lasted fewer than 80 100–109
hours?
(3) Which interval contains the greatest number of batteries?
(4) Which interval contains the least number of batteries?
c. Display the data in a stem-and-leaf diagram. Use the digits from 5 through 10 as the
stems.
d. What is the range of the data?
e. What is the probability that a battery selected at random lasted more than 100 hours?
5. The following data consist of the hours spent each week watching television, as reported by
a group of 38 teenagers:
13, 20, 17, 36, 25, 21, 9, 32, 20, 17, 12, 19, 5, 8, 11, 28, 25, 18,
19, 22, 4, 6, 0, 10, 16, 3, 27, 31, 15, 18, 20, 17, 3, 6, 19, 25, 4, 7
a. Construct a table to group these data, using intervals of 0–4, 5–9, 10–14, 15–19, 20–24,
25–29, 30–34, and 35–39.
b. Construct a table to group these data, using intervals of 0–7, 8–15, 16–23, 24–31, and
32–39.
c. Display the data in a stem-and-leaf diagram.
d. What is the range of the data?
e. What is the probability that a teenager, selected at random from this group, spends less
than 4 hours watching television each week?
6. The following data show test scores for 30 students:
90, 83, 87, 71, 62, 46, 67, 72, 75, 100, 93, 81, 74, 75, 82,
83, 83, 84, 92, 58, 95, 98, 81, 88, 72, 59, 95, 50, 73, 93
674 Statistics

a. Construct a table, using intervals of length 10 starting with 91–100.

b. Construct a table, using intervals of length 12 starting with 89–100.
c. For the grouped data in part a, which interval contains the greatest number of stu-
dents?
d. For the grouped data in part b, which interval contains the greatest number of stu-
dents?
e. Do the answers for parts c and d indicate the same general region of test scores, such
as “scores in the eighties”? Explain your answer.
7. For the ungrouped data from Exercise 5, tell why each of the following sets of intervals is
not correct for grouping the data.
a. b. c. d.
Interval Interval Interval Interval
25–38 30–39 32–40 33–40
13–24 20–29 24–32 25–32
0–12 10–19 16–24 17–24
5–9 8–16 9–16
0–4 0–8 1–8

Hands-On Activity
Organize the data that you collected in the Hands-On Activity for Section 16-1.
1. Use a stem-and-leaf diagram.
a. Decide what will be used as stems.
b. Decide what will be used as leaves.
c. Construct the diagram.
d. Check that the number of leaves in the diagram equals the number of values in the
data collected.
2. Use a frequency table.
a. How many intervals will be used?
b. What will be the length of each interval?
c. What will be the starting and ending points of each interval? Check that the intervals
do not overlap, are equal in size, and that every value falls into only one interval.
d. Tally the data.
e. List the frequency for each interval.
f. Check that the total frequency equals the number of values in the data collected.
3. Decide which method of organization is better for your data. Explain your choice.

Keep your organized data to work with as you learn more about statistics.
 The Histogram 675

16-3 THE HISTOGRAM

In Section 16-2 we organized data by grouping them into intervals of equal
length. After the data have been organized, a graph can be used to visualize the
intervals and their frequencies.
The table below shows the distribution of test scores for 32 students in a
class. The data have been organized into six intervals of length 10.

Test Scores (Intervals) Frequency (Number of Scores)

91–100 6
81–90 8
71–80 11
61–70 4
51–60 0
41–50 3

We can use a histogram to display the data graphically. A histogram is a ver-

tical bar graph in which each interval is represented by the width of the bar and
the frequency of the interval is represented by the height of the bar. The bars
are placed next to each other to show that, as one interval ends, the next inter-
val begins.
TEST SCORES OF 32 STUDENTS
12
11
10
9
8
Frequency

7
6
5
4
3
2
1
0
41–50 51–60 61–70 71–80 81–90 91–100
Test scores (intervals)
In the above histogram, the intervals are listed on the horizontal axis in the
order of increasing scores, and the frequency scale is shown on the vertical axis.
The first bar shows that 3 students had test scores in the interval 41–50. Since no
student scored in the interval 51–60, there is no bar for this interval. Then, 4 stu-
dents scored between 61 and 70; 11 between 71 and 80; 8 between 81 and 90; and
6 between 91 and 100.
676 Statistics

Except for an interval having a frequency of 0, the interval 51–60 in this

example, there are no gaps between the bars drawn in a histogram. Since the his-
togram displays the frequency, or number of data values, in each interval, we
sometimes call this graph a frequency histogram.

A graphing calculator can display a frequency histogram from the data on a

frequency distribution table.
(1) Clear L1 and L2 with the ClrList function by pressing STAT 4 2nd

L1 , 2nd L2 ENTER .

(2) Press STAT 1 to edit the lists. L1 will L1 L2 L3 2

contain the minimum value of each 91 6 ------
81 8
interval. Move the cursor to the first entry 71 11
position in L1. Type the value and then 61 4
51 0
press ENTER . Type the next value and 41 3
------ ------
then press ENTER . Repeat this process
until all the minimum values of the L2(7) =
intervals have been entered.
(3) Repeat the process to enter the frequencies that correspond to each inter-
val in L2.
(4) Clear any functions in the Y= menu.
(5) Turn on Plot1 from the STAT PLOT
Plot1 Plot2 Plot3
menu, and configure it to graph a his-
togram. Make sure to also set Xlist to L1 On = Off
and Freq to L2. Ty p e :
ENTER: 2nd STAT PLOT 1 ENTER Xlist: L1
 佡 佡 ENTER 佡 2nd Freq: L2
L1  2nd L2

(6) In the WINDOW menu, accessed by

Window
pressing WINDOW , enter Xmin as 31, Xmin=31
the length of one interval less than the Xmax=110
Xsc1=10
smallest interval value and Xmax as 110, Ymin=0
the length of one interval more than the Ymax=12
largest interval value. Enter Xscl as 10, Ysc1=1
Xres=1
the length of the interval. The Ymin is 0
and Ymax is 12 to be greater than the
largest frequency.
 The Histogram 677

(7) Press GRAPH to draw the graph. We can P1:L1,L2

view the frequency (n) associated with
each interval by pressing TRACE . Use the
left and right arrow keys to move between
intervals.
min=81
max<91 n=8

EXAMPLE 1

The table on the right represents the number of

Interval Frequency
miles per gallon of gasoline obtained by 40 drivers
of compact cars. Construct a frequency histogram 16–19 5
based on the data.
20–23 11
Solution (1) Draw and label a vertical scale to show fre- 24–27 8
quencies. The scale starts at 0 and increases to 28–31 5
include the highest frequency in any one inter-
32–35 7
val (here, it is 11).
36–39 3
(2) Draw and label intervals of equal length on a 40–43 1
horizontal scale. Label the horizontal scale,
telling what the numbers represent.

(3) Draw the bars vertically, leaving no gaps between the intervals.

12
11
10
9
8
Frequency

7
6
5
4
3
2
1
0
16–19 20–23 24–27 28–31 32–35 36–39 40–43
Mileage (miles per gallon) for compact cars
678 Statistics

Calculator (1) Press STAT 1 to edit the lists and enter the minimum value of each
Solution interval into L1: 16, 20, 24, 28, 32, 36, 40. Use the arrow key to move into
L2, and enter the corresponding frequencies: 5, 11, 8, 5, 7, 3, 1.
(2) Go to the STAT PLOT menu and choose Plot1 by pressing 2nd
STAT PLOT 1 . Move the cursor with the arrow keys, then press
ENTER to select On and the histogram. Type 2nd L1 into Xlist and
2nd L2 into Freq.
(3) Set the Window. Each interval has length 4, so set Xmin to 12 (4 less than
the smallest interval value), Xmax to 44 (4 more than the largest interval
value), and Xscl to 4. Make Ymin 0 and Ymax 12 to be greater than the
largest frequency.
P1:L1,L2
(4) Draw the graph by pressing GRAPH . Press
TRACE and use the right and left arrow
keys to show the frequencies, the heights of
the vertical bars.

min=16
max<20 n=5

EXAMPLE 2

Use the histogram constructed in Example 1 to answer the following questions:

a. In what interval is the greatest frequency found?
b. What is the number (or frequency) of cars reporting mileages between 28
and 31 miles per gallon?
c. For what interval are the fewest cars reported?
d. How many of the cars reported mileage greater than 31 miles per gallon?
e. What percent of the cars reported mileage from 24 to 27 miles per gallon?

Solution a. 20–23
b. 5
c. 40–43
d. Add the frequencies for the three highest intervals. The interval 32–35 has a
frequency of 7; 36–39 a frequency of 3; 40–43 a frequency of 1: 7  3  1  11.
e. The interval 24–27 has a frequency of 8. The total frequency for this survey
8
is 40. 40 5 15  20%.

Answers a. 20–23 b. 5 c. 40–43 d. 11 e. 20%

 The Histogram 679

EXERCISES
Writing About Mathematics
1. Compare a stem-and-leaf diagram with a frequency histogram. In what ways are they alike
and in what ways are they different?
2. If the data in Example 1 had been grouped into intervals with a lowest interval of 16–20,
what would be the endpoints for the other intervals? Would you be able to determine the
frequency for each new interval? Explain why or why not.

Developing Skills
In 3–5, in each case, construct a frequency histogram for the grouped data. Use graph paper or a
graphing calculator.
3. 4. 5.
Interval Frequency Interval Frequency Interval Frequency

91–100 5 30–34 5 1–3 24

81–90 9 25–29 10 4–6 30
71–80 7 20–24 10 7–9 28
61–70 2 15–19 12 10–12 41
51–60 4 10–14 0 13–15 19
5–9 2 16–18 8

6. For the table of grouped data given in Exercise 5, answer the following questions:
a. What is the total frequency in the table?
b. What interval contains the greatest frequency?
c. The number of data values reported for the interval 4–6 is what percent of the total
number of data values?
d. How many data values from 10 through 18 were reported?

Applying Skills
7. Towering Ted McGurn is the star of the school’s basketball team. The number of points
scored by Ted in his last 20 games are as follows:
36, 32, 28, 30, 33, 36, 24, 33, 29, 30, 30, 25, 34, 36, 34, 31, 36, 29, 30, 34
a. Copy and complete the table to find the frequency for
Interval Tally Frequency
each interval.
b. Construct a frequency histogram based on the data 35–37
found in part a. 32–34
c. Which interval contains the greatest frequency? 29–31
d. In how many games did Ted score 32 or more points? 26–28
e. In what percent of these 20 games did Ted score fewer 23–25
than 26 points?
680 Statistics

8. Thirty students on the track team were timed in the 200-meter dash. Each student’s time
was recorded to the nearest tenth of a second. Their times are as follows:
29.3, 31.2, 28.5, 37.6, 30.9, 26.0, 32.4, 31.8, 36.6, 35.0,
38.0, 37.0, 22.8, 35.2, 35.8, 37.7, 38.1, 34.0, 34.1, 28.8,
29.6, 26.9, 36.9, 39.6, 29.9, 30.0, 36.0, 36.1, 38.2, 37.8
a. Copy and complete the table to find the frequency in
Interval Tally Frequency
each interval.
b. Construct a frequency histogram for the given data. 37.0–40.9
c. Determine the number of students who ran the 200- 33.0–36.9
meter dash in under 29 seconds. 29.0–32.9
d. If a student on the track team is chosen at random, 25.0–28.9
what is the probability that he or she ran the 200-
21.0–24.9
meter dash in fewer than 29 seconds?

Hands-On-Activity
Construct a histogram to display the data that you collected and organized in the Hands-On
Activities for Sections 16-1 and 16-2.
1. Draw the histogram on graph paper.
2. Follow the steps in this section to display the histogram on a graphing calculator.

16-4 THE MEAN, THE MEDIAN, AND THE MODE

In a statistical study, after we have collected the data, organized them, and pre-
sented them graphically, we then analyze the data and summarize our findings.
To do this, we often look for a representative, or typical, score.

Averages in Arithmetic
In your previous study of arithmetic, you learned how to find the average of two
or more numbers. For example, to find the average of 17, 25, and 30:
STEP 1. Add these three numbers: 17  25  30  72.
STEP 2. Divide this sum by 3 since there are three numbers: 72  3  24.
The average of the three numbers is 24.

Averages in Statistics
The word average has many different meanings. For example, there is an aver-
age of test scores, a batting average, the average television viewer, an average
intelligence, and the average size of a family. These averages are not found by
 The Mean, the Median, and the Mode 681

the same rule or procedure. Because of this confusion, in statistics we speak of

measures of central tendency. These measures are numbers that usually fall
somewhere in the center of a set of organized data.
We will discuss three measures of central tendency: the mean, the median,
and the mode.

The Mean
In statistics, the arithmetic average previously studied is called the mean of a set
of numbers. It is also called the arithmetic mean or the numerical average. The
mean is found in the same way as the arithmetic average is found.

Procedure
To find the mean of a set of n numbers, add the numbers and divide
the sum by n.The symbol used for the mean is x–.

For example, if Ralph’s grades on five tests in science during this marking
period are 93, 80, 86, 72, and 94, he can find the mean of his test grades as fol-
lows:
STEP 1. Add the five data values: 93  80  86  72  94  425.
STEP 2. Divide this sum by 5, the number of tests: 425  5  85.
The mean (arithmetic average) is 85.
Let us consider another example. In a car wash, there are seven employees
whose ages are 17, 19, 20, 17, 46, 17, and 18. What is the mean of the ages of these
employees?
Here, we add the seven ages to get a sum of 154. Then, 154  7  22. While
the mean age of 22 is the correct answer, this measure does not truly represent
the data. Only one person is older than 22, while six people are under 22. For
this reason, we will look at another measure of central tendency that will elimi-
nate the extreme case (the employee aged 46) that is distorting the data.

The Median
The median is the middle value for a set of data arranged in numerical order.
For example, the median of the ages 17, 19, 20, 17, 46, 17, and 18 for the car-wash
employees can be found in the following manner:
STEP 1. Arrange the ages in numerical order: 17, 17, 17, 18, 19, 20, 46
STEP 2. Find the middle number: 17, 17, 17, 18, 19, 20, 46
↑
The median is 18 because there are three ages less than 18 and three ages
greater than 18. The median, 18, is a better indication of the typical age of the
682 Statistics

employees than the mean, 22, because there are so many younger people work-
ing at the car wash.
Now, let us suppose that one of the car-wash employees has a birthday, and
her age changes from 17 to 18. What is now the median age?
STEP 1. Arrange the ages in numerical order: 17, 17, 18, 18, 19, 20, 46
STEP 2. Find the middle number: 17, 17, 18, 18, 19, 20, 46
↑
The median, or middle value, is again 18. We can no longer say that there are
three ages less than 18 because one of the three youngest employees is now 18.
We can say, however, that:
1. the median is 18 because there are three ages less than or equal to 18 and
three ages greater than or equal to 18; or
2. the median is 18 because, when the data values are arranged in numerical
order, there are three values below this median, or middle number, and
three values above it.
Recently, the car wash hired a new employee whose age is 21. The data now
include eight ages, an even number, so there is no middle value. What is now the
median age?
STEP 1. Arrange the ages in numerical order: 17, 17, 18, 18, 19, 20, 21, 46
STEP 2. There is no single middle number. 17, 17, 18, 18, 19, 20, 21, 46
Find the two middle numbers: ↑ ↑
18 1 19
STEP 3. Find the mean (arithmetic average) 2 5 1812
of the two middle numbers:
The median is now 1812. There are four ages less than this center value of 1812
and four ages greater than 1812.

Procedure
To find the median of a set of n numbers:
1. Arrange the numbers in numerical order.
2. If n is odd, find the middle number.This number is the median.
3. If n is even, find the mean (arithmetic average) of the two middle numbers.
This average is the median.

The Mode
The mode is the data value that appears most often in a given set of data. It is
usually best to arrange the data in numerical order before finding the mode.
 The Mean, the Median, and the Mode 683

Let us consider some examples of finding the mode:

1. The ages of employees in a car wash are 17, 17, 17, 18, 19, 20, 46. The
mode, which is the number appearing most often, is 17.
2. The number of hours each of six students spent reading a book are
6, 6, 8, 11, 14, 21. The mode, or number appearing most frequently, is 6.
In this case, however, the mode is not a useful measure of central ten-
dency. A better indication is given by the mean or the median.
3. The number of photographs printed from each of Renee’s last six rolls of
film are 8, 8, 9, 11, 11, and 12. Since 8 appears twice and 11 appears twice,
we say that there are two modes: 8 and 11. We do not take the average of
these two numbers since the mode tells us where most of the scores
appear. We simply report both numbers. When two modes appear within a
set of data, we say that the data are bimodal.
4. The number of people living in each house on Meryl’s street are 2, 2, 3, 3,
4, 5, 5, 6, 8. These data have three modes: 2, 3, and 5.
5. Ralph’s test scores in science are 72, 80, 86, 93, and 94. Here, every number
appears the same number of times, once. Since no number appears more
often than the others, we define such data as having no mode.

Procedure
To find the mode for a set of data, find the number or numbers that
occur most often.
1. If one number appears most often in the data, that number is the mode.
2. If two or more numbers appear more often than all other data values, and
these numbers appear with the same frequency, then each of these numbers
is a mode.
3. If each number in a set of data occurs with the same frequency, there is no
mode.

KEEP IN MIND Three measures of central tendency are:

1. The mean, or mean average, found by adding n data values and then divid-
ing the sum by n.
2. The median, or middle score, found when the data are arranged in numeri-
cal order.
3. The mode, or the value that appears most often.

A graphing calculator can be used to arrange the data in numerical order

and to find the mean and the median. The calculator solution in the following
example lists the keystrokes needed to do this.
684 Statistics

EXAMPLE 1

The weights, in pounds, of five players on the basketball team are 195, 168, 174,
182, and 181. Find the average weight of a player on this team.

Solution The word average, by itself, indicates the mean. Therefore:

(1) Add the five weights: 195  168  174  182  181  900.
(2) Divide the sum by 5, the number of players: 900  5  180.

Calculator Enter the data into list L1. Then use 1-Var Stats from the STAT CALC menu
Solution to display information about this set of data.

ENTER: STAT 佡 ENTER ENTER

DISPLAY:
1–Var Stats
–
x=180
x=900
x2=162410
Sx=10.12422837
x=9.055385138
In=5
<

The first value given is 2

x , the mean.

Answer 180 pounds

The second value given is Σx  900. The symbol Σ represents a sum and
Σx  900 can be read as “The sum of the values of x is 900.” The list shows other
values related to this set of data. The arrow at the bottom of the display indi-
cates that more entries follow what appears on the screen. These can be dis-
played by pressing the down arrow. One of these is the median (Med  181).
The display also shows that there are 5 data values (n = 5). Others we will use
in later sections in this chapter and in more advanced courses.

EXAMPLE 2

Renaldo has marks of 75, 82, and 90 on three mathematics tests. What mark
must he obtain on the next test to have an average of exactly 85 for the four
math tests?
 The Mean, the Median, and the Mode 685

Solution The word average, by itself, indicates the mean.

Let x  Renaldo’s mark on the fourth test.
The sum of the four test marks divided by 4 is 85. Check
75 1 82 1 90 1 x 75 1 82 1 90 1 93 ?
4 5 85 4 5 85
247 1 x 340 ?
4 5 85 4 5 85
?
247 1 x 5 340 85 5 85 ✔
x 5 93

Answer Renaldo must obtain a mark of 93 on his fourth math test.

EXAMPLE 3

Find the median for each distribution.

a. 4, 2, 5, 5, 1 b. 9, 8, 8, 7, 4, 3, 3, 2, 0, 0

Solution a. Arrange the data in numerical order: 1, 2, 4, 5, 5

The median is the middle value: 1, 2, 4, 5, 5
↑

Answer median  4
b. Since there is an even number of values, there are two middle values. Find
the mean (average) of these two middle values:
9, 8, 8, 7, 4, 3, 3, 2, 0, 0
↑ ↑
413
2 5 72 5 312

Answer median  312 or 3.5

EXAMPLE 4

Find the mode for each distribution.

a. 2, 9, 3, 7, 3 b. 3, 4, 5, 4, 3, 7, 2 c. 1, 2, 3, 4, 5, 6, 7

Solution a. Arrange the data in numerical order: 2, 3, 3, 7, 9.

The mode, or most frequent value, is 3.
b. Arrange the data in numerical order: 2, 3, 3, 4, 4, 5, 7. Both 3 and 4 appear
twice. There are two modes.
c. Every value occurs the same number of times in the data set 1, 2, 3, 4, 5, 6, 7.
There is no mode.

Answers a. The mode is 3. b. The modes are 3 and 4. c. There is no mode.

686 Statistics

Linear Transformations of Data

Multiplying each data value by the same constant or adding the same constant
to each data value is an example of a linear transformation of a set of data.
Let us start by examining additive transformations. For instance, consider
the data 2, 2, 3, 4, 5. If 10 is added to each data value, the data set becomes:
12, 12, 13, 14, 15
Notice that every measure of central tendency has been shifted to the right by
10 units:
old mean  2 1 2 1 53 1 4 1 5 5 3.2 new mean  12 1 12 1 13
5
1 14 1 15
5 13.2
old median  3 new median  13
old mode  2 new mode  12
In fact, this result is valid for any additive transformation of a data set. In
general:
 If x–, d, and o are the mean, median, and mode of a set of data and the con-
stant c is added to each data value, then x–  c, d  c, and o  c are the mean,
median, and mode of the transformed data.

It can be also shown that a similar result holds for multiplicative transfor-
mations, that is:
 If x–, d, and o are the mean, median, and mode of a set of data and each data
value is multiplied by the nonzero constant c, then cx–, cd, and co are the
mean, median, and mode of the transformed data.

EXAMPLE 5

In Ms. Huan’s Algebra class, the average score on the most recent quiz was 65.
Being in a generous mood, Ms. Huan decided to curve the quiz by adding 10
points to each quiz score. What will be the new average score for the class?

Answer 65  10  75 points

EXERCISES
Writing About Mathematics
1. On her first two math tests, Rene received grades of 67 and 79. Her mean (average) grade
for these two tests was 73. On her third test she received a grade of 91. Rene found the
mean of 73 and 91 and said that her mean for the three tests was 82. Do you agree with
Rene? Explain why or why not.
 The Mean, the Median, and the Mode 687

2. Carlos said that when there are n numbers in a set of data and n is an odd number, the
median is the n 1 1
2 th number when the data are arranged in order. Do you agree with
Carlos? Explain why or why not.

Developing Skills
3. For each set of data, find the mean.
a. 7, 3, 5, 11, 9 b. 22, 38, 18, 14, 22, 30
c. 512, 234, 712, 534, 412 d. 1.00, 0.01, 1.10, 0.12, 1.00, 1.03
4. Find the median for each set of data.
a. 1, 2, 5, 3, 4 b. 2, 9, 2, 9, 7
c. 3, 8, 12, 7, 1, 0, 4 d. 80, 83, 97, 79, 25
e. 3.2, 8.7, 1.4 f. 2.00, 0.20, 2.20, 0.02, 2.02
g. 21, 24, 23, 22, 20, 24, 23, 21, 22, 23 h. 5, 7, 9, 3, 8, 7, 5, 6
5. What is the median for the digits 1, 2, 3, . . . , 9?
6. What is the median for the counting numbers from 1 through 100?
7. Find the mode for each distribution.
a. 2, 2, 3, 4, 8 b. 2, 2, 3, 8, 8
c. 2, 2, 8, 8, 8 d. 2, 3, 4, 7, 8
e. 2, 2, 3, 8, 8, 9, 9 f. 1, 2, 1, 2, 1, 2, 1
g. 1, 2, 3, 2, 1, 2, 3, 2, 1 h. 3, 19, 21, 75, 0, 6
i. 3, 2, 7, 6, 2, 7, 3, 1, 4, 2, 7, 5 j. 19, 21, 18, 23, 19, 22, 18, 19, 20
8. A set of data consists of six numbers: 7, 8, 8, 9, 9, and x. Find the mode for these six numbers
when:
a. x  9 b. x  8 c. x  7 d. x  6
9. A set of data consists of the values 2, 4, 5, x, 5, 4. Find a possible value of x such that:
a. there is no mode because all scores appear an equal number of times
b. there is only one mode
c. there are two modes
10. For the set of data 5, 5, 6, 7, 7, which statement is true?
(1) mean  mode (3) mean  median
(2) median  mode (4) mean  median
11. For the set of data 8, 8, 9, 10, 15, which statement is true?
(1) mean  median (3) median  mode
(2) mean  mode (4) mean  median
688 Statistics

12. When the data consists of 3, 4, 5, 4, 3, 4, 5, which statement is true?

(1) mean  median (3) median  mode
(2) mean  mode (4) mean  median
13. For which set of data is there no mode?
(1) 2, 1, 3, 1, 2 (3) 1, 2, 4, 3, 5
(2) 1, 2, 3, 3, 3 (4) 2, 2, 3, 3, 3
14. For which set of data is there more than one mode?
(1) 8, 7, 7, 8, 7 (3) 8, 7, 5, 7, 6, 5
(2) 8, 7, 4, 5, 6 (4) 1, 2, 2, 3, 3, 3
15. For which set of data does the median equal the mode?
(1) 3, 3, 4, 5, 6 (3) 3, 3, 4
(2) 3, 3, 4, 5 (4) 3, 4
16. For which set of data will the mean, median, and mode all be equal?
(1) 1, 2, 5, 5, 7 (3) 1, 1, 1, 2, 5
(2) 1, 2, 5, 5, 8, 9 (4) 1, 1, 2
17. The median of the following data is 11:
2, 5, 9, 11, 40, 3, 4, 5, 10, 45, 32, 40, 67, 7, 11, 9, 20, 34, 5, 1, 8, 15, 16, 19, 39
a. If 4 is subtracted from each data value, what is the median of the transformed data set?
b. If the largest data value is doubled and the smallest data value is halved, what is the
median of the new data set?
18. The mean of the following data is 37.625:
3, 0, 1, 7, 8, 11, 31, 15, 99, 98, 92, 81, 85, 87, 55, 54, 34, 27, 26, 21, 14, 17, 19, 18
If each data value is multiplied by 2 and increased by 5, what is the mean of the trans-
formed data set?
19. Three consecutive integers can be represented by x, x  1, and x  2. The average of these
consecutive integers is 32. What are the three integers?
20. Three consecutive even integers can be represented by x, x  2, and x  4. The average of
these consecutive even integers is 20. Find the integers.
21. The mean of three numbers is 31. The second is 1 more than twice the first. The third is 4
less than 3 times the first. Find the numbers.

Applying Skills
22. Sid received grades of 92, 84, and 70 on three tests. Find his test average.
23. Sarah’s grades were 80 on each of two of her tests and 90 on each of three other tests. Find
her test average.
24. Louise received a grade of x on each of two of her tests and of y on each of three other
tests. Represent her average for all the tests in terms of x and y.
 The Mean, the Median, and the Mode 689

25. Andy has grades of 84, 65, and 76 on three social studies tests. What grade must he obtain
on the next test to have an average of exactly 80 for the four tests?
26. Rosemary has grades of 90, 90, 92, and 78 on four English tests. What grade must she obtain
on the next test so that her average for the five tests will be 90?
27. The first three test scores are shown below for each of four students. A fourth test will be
given and averages taken for all four tests. Each student hopes to maintain an average of 85.
Find the score needed by each student on the fourth test to have an 85 average, or explain
why such an average is not possible.
a. Pat: 78, 80, 100 b. Bernice: 79, 80, 81
c. Helen: 90, 92, 95 d. Al: 65, 80, 80
28. The average weight of Sue, Pam, and Nancy is 55 kilograms.
a. What is the total weight of the three girls?
b. Agnes weighs 60 kilograms. What is the average weight of the four girls: Sue, Pam,
Nancy, and Agnes?
29. For the first 6 days of a week, the average rainfall in Chicago was 1.2 inches. On the last day
of the week, 1.9 inches of rain fell. What was the average rainfall for the week?
30. If the heights, in centimeters, of a group of students are 180, 180, 173, 170, and 167, what is
the mean height of these students?
31. What is the median age of a family whose members are 42, 38, 14, 13, 10, and 8 years old?
32. What is the median age of a class in which 14 students are 14 years old and 16 students are
15 years old?
33. In a charity collection, ten people gave amounts of $1, $2, $1, $1, $3, $1, $2, $1, $1, and $1.50.
What was the median donation?
34. The test scores for an examination were 62, 67, 67, 70, 90, 93, and 98. What is the median test
score?
35. The weekly salaries of six employees in a small firm are $440, $445, $445, $450, $450, and $620.
a. For these six salaries, find: (1) the mean (2) the median (3) the mode
b. If negotiations for new salaries are in session and you represent management, which
measure of central tendency will you use as the average salary? Explain your answer.
c. If negotiations are in session and you represent the labor union, which measure of cen-
tral tendency will you use as an average salary? Explain your answer.
36. In a certain school district, bus service is provided for students living at least 112 miles from
school. The distances, rounded to the nearest half mile, from school to home for ten students
are 0, 21, 12, 1, 1, 1, 1, 112, 312, and 10 miles.
a. For these data, find: (1) the mean (2) the median (3) the mode
b. How many of the ten students are entitled to bus service?
c. Explain why the mean is not a good measure of central tendency to describe the aver-
age distance between home and school for these students.
690 Statistics

37. Last month, a carpenter used 12 boxes of nails each of which contained nails of only one
size. The sizes marked on the boxes were:
3
4 in., 34 in., 34 in., 34 in., 34 in., 34 in., 34 in., 34 in., 1 in., 1 in., 2 in., 2 in.
a. For these data, find: (1) the mean (2) the median (3) the mode
b. Describe the average-size nail used by the carpenter, using at least one of these mea-
sures of central tendency. Explain your answer.

Hands-On Activity
Find the mean, the median, and the mode for the data that you collected in the Hands-On Activity
for Section 16-1. It may be necessary to go back to your original data to do this.

16-5 MEASURES OF CENTRAL TENDENCY AND GROUPED DATA

Intervals of Length 1
In a statistical study, when the range is small, we can use intervals of length 1 to
group the data. For example, each member of a class of 25 students reported the
number of books he or she read during the first half of the school year. The data
are as follows:
5, 3, 5, 3, 1, 8, 2, 4, 2, 6, 3, 8, 8, 5, 3, 4, 5, 8, 5, 3, 3, 5, 6, 2, 3
These data, for which the values range from 1 to
Interval Frequency
8, can be organized into a table such as the one
shown at the right, with each value representing 8 4
an interval.
7 0
Since 25 students were included in this study,
the total frequency, N, is 25. We can use this 6 2
table, with intervals of length 1, to find the mode, 5 6
median, and mean for these data.
4 2
3 7
2 3
1 1
N  25

Mode of a Set of Grouped Data

Since the greatest frequency, 7, appears for interval 3, the mode for the data is
3. In general:
 For a set of grouped data, the mode is the value of the interval that contains
the greatest frequency.
 Measures of Central Tendency and Grouped Data 691

Median of a Set of Grouped Data

We have learned that the median for a set of data in numerical order is the mid-
dle value.
For these 25 numbers, there are 12 numbers greater than or equal to the
median, and 12 numbers less than or equal to the median. Therefore, when the
numbers are written in numerical order, the median is the 13th number from
either end.

1, 2, 2, 2, 3, 3, 3, 3, 3, 3, 3, 4, 4, 5, 5, 5, 5, 5, 5, 6, 6, 8, 8, 8, 8
↑– The median is 4.

When the data are grouped in the table shown earlier, a simple counting
procedure can be used to find the median, the 13th number. When we add the
frequencies of the first four intervals, starting at the top, we find that these inter-
vals include data for:

4  0  2  6  12 students

Therefore, the next lower interval (with frequency greater than 0) must in-
clude the median, the value for the 13th student. This is the interval for the data
value 4.
When we add the frequencies of the first three intervals, starting at the bot-
tom, we find that these intervals include data for:

1  3  7  11 students

The next higher interval contains two scores, one for the 12th student and one
that is the median, or the value for the 13th student. Again this is the interval for
the data value 4.
In general:

 For a set of grouped data, the median is the value of the interval that con-
tains the middle data value.

Mean of a Set of Grouped Data

By adding the four 8’s in the ungrouped data, we see that four students, read-
ing eight books each, have read 8  8  8  8 or 32 books. We can arrive at
this same number by using the grouped intervals in the table: we multiply the
four 8’s by the frequency 4. Thus, (4)(8) 5 32. Applying this multiplication
shortcut to each row of the table, we obtain the third column of the following
table:
692 Statistics

Interval Frequency (Interval)  (Frequency)

8 4 8  4  32
7 0 700
6 2 6  2  12
5 6 5  6  30
4 2 428
3 7 3  7  21
2 3 236
1 1 111
N = 25 Total  110

The total (110) represents the sum of all 25 pieces of data. We can check this
by adding the 25 scores in the unorganized data.
Finally, to find the mean, we divide the total number, 110, by the number of
items, 25. Thus, the mean for the data is: 110  25  4.4.

Procedure
To find the mean for N values in a table of grouped data when the
length of each interval is 1:
1. For each interval, multiply the interval value by its corresponding frequency.
2. Find the sum of these products.
3. Divide this sum by the total frequency, N.

Calculator Solution for Grouped Data

The calculator can be used to find the mean and median for the grouped data
shown above. Enter the number of books read by each student into L1 and the
frequency for each number of books into L2. Then use the 1-Var Stats from the
STAT CALC menu to display information about the data.
ENTER: STAT 佡 ENTER 2nd L1 , 2nd L2 ENTER

DISPLAY:
1–Var Stats
–
x=4.4
x=110
x2=586
Sx=2.061552813
x=2.019900988
In=25
<
 Measures of Central Tendency and Grouped Data 693

The display shows that the mean, x–, is 4.4, the sum of the number of books read
is 110, and the number of students, the total frequency, N, is 25. Use the down
arrow to display the median, Med  4.

Intervals Other Than Length 1

There are specific mathematical procedures to find the mean, median, and
mode for grouped data with intervals other than length 1, but we will not study
them at this time. Instead, we will simply identify the intervals that contain some
of these measures of central tendency. For example, a small industrial plant sur-
veyed 50 workers to find the number of miles each person commuted to work.
The commuting distances were reported, to the nearest mile, as follows:
0, 0, 1, 1, 2, 2, 2, 3, 3, 4, 4, 4, 5, 5, 6, 6, 6, 7, 7, 7, 9,
10, 10, 10, 10, 10, 10, 10, 10, 12, 12, 14, 15, 17, 17,
18, 22, 23, 25, 28, 30, 32, 32, 33, 34, 34, 36, 37, 37, 52
These data are organized into a table with intervals of length 10, as follows:

Interval Frequency
(commuting distance) (number of workers)

50–59 1
40–49 0
30–39 9
20–29 4
10–19 15
0–9 21
N  50

Modal Interval
In the table, interval 0–9 contains the greatest frequency, 21. We say that inter-
val 0–9 is the group mode, or modal interval, because this group of numbers has
the greatest frequency. The modal interval is not the same as the mode. The
modal interval is a group of numbers; the mode is usually a single number. For
this example, the original data (before being placed into the table) show that the
number appearing most often is 10. Hence, the mode is 10. The modal interval,
which is 0–9, tells us that, of the six intervals in the table, the most frequently
occurring commuting distance is 0 to 9 miles.
Both the mode and the modal interval depend on the concept of greatest
frequency. For the mode, we look for a single number that has the greatest fre-
quency. For the modal interval, we look for the interval that has the greatest
frequency.
694 Statistics

Interval Containing the Median

To find the interval containing the median, we follow the procedure described
earlier in this section. For 50 numbers, the median, or middle number, will be at
a point where 25 numbers are at or above the median and 25 are at or below it.
Count the frequencies in the table from the uppermost interval and move
downward. We add 1  0  9  4  14. Since there are 15 numbers in the next
lower interval, and 14 1 15 5 29, we see that the 25th number will be reached
somewhere in that interval, 10–19.
Count from the bottom interval and move up. We have 21 numbers in
the first interval. Since there are 15 numbers in the next higher interval, and
21  15  36, we see that 25th number will be reached somewhere in that inter-
val, 10–19. This is the same result that we obtained when we moved downward.
The interval containing the median for this grouping is 10–19.
In this course, we will not deal with problems in which the median is not
found in any interval.

Interval Containing the Mean

When data are grouped using intervals of length other than 1, there is no sim-
ple procedure to identify the interval containing the mean. However, the mean
can be approximated by assuming that the data are equally distributed through-
out each interval. The mean is then found by using the midpoint of each inter-
val as the value of each entry in the interval. This problem is studied in
higher-level courses.

EXAMPLE 1

In the table, the data indicate the heights, in inches,

Height Frequency
of 17 basketball players. For these data find:
(inches) (number)
a. the mode b. the median c. the mean
77 2
Solution a. The greatest frequency, 5, occurs for the height
76 0
of 75 inches. The mode, or height appearing
most often, is 75. 75 5

b. For 17 players, the median is the 9th number, so 74 3

there are 8 heights greater than or equal to the 73 4
median and 8 heights less than or equal to the 72 2
median. Counting the frequencies going down,
we have 2  0  5  7. Since the frequency of 71 1
the next interval is 3, the 8th, 9th, and 10th
heights are in this interval, 74.
Counting the frequencies going up, we have 1  2  4  7. Again, the fre-
quency of the next interval is 3, and the 8th, 9th, and 10th heights are in this
interval. The 9th height, the median, is 74.
 Measures of Central Tendency and Grouped Data 695

c. (1) Multiply each height by its corresponding frequency:

77  2  154 76  0  0 75  5  375 74  3  222
73  4  292 72  2  144 71  1  71
(2) Find the total of these products:
154  0  375  222  292  144  71  1,258
(3) Divide this total, 1,258, by the total frequency, 17 to obtain the mean:
1258  17  74

Calculator Clear any previous data that may be stored in L1 and L2. Enter the heights of
Solution the players into L1 and the frequencies into L2. Then use 1-Var Stats from the
STAT CALC menu to display information about the data. The screen will
show the mean, x–. Press the down arrow key to display the median.

ENTER: STAT 佡 ENTER 2nd L1 , 2nd L2 ENTER

DISPLAY:
1–Var Stats 1–Var Stats
–

<
I
x=74 n=17
x=1258 MinX=71
x2=93136 Q1=73
Sx=1.658312395 Med=74
x=1.608799333 Q 3= 7 5
In=17
<

MaxX=77

Answers a. mode  75 b. median  74 c. mean  74

EXERCISES
Writing About Mathematics
1. The median for a set of 50 data values is the average of the 25th and 26th data values when
the data is in numerical order. What must be true if the median is equal to one of the data
values? Explain your answer.
2. What must be true about a set of data if the median is not one of the data values? Explain
your answer.
696 Statistics

Developing Skills
In 3–5, the data are grouped in each table in intervals of length 1.
Find: a. the total frequency b. the mean c. the median d. the mode

3. 4. 5.
Interval Frequency Interval Frequency Interval Frequency

10 1 15 3 25 4
9 2 16 2 24 0
8 3 17 4 23 3
7 3 18 1 22 2
6 4 19 5 21 4
5 3 20 6 20 5
19 2

In 6–8, the data are grouped in each table in intervals other than length 1. Find: a. the total frequency
b. the interval that contains the median c. the modal interval

6. 7. 8.
Interval Frequency Interval Frequency Interval Frequency

55–64 3 4–9 12 126–150 4

45–54 8 10–15 13 101–125 6
35–44 7 16–21 9 76–100 6
25–34 6 22–27 12 51–75 3
15–24 2 28–33 15 26–50 7
34–39 10 1–25 2

Applying Skills
9. On a test consisting of 20 questions, 15 students received the following scores:

17, 14, 16, 18, 17, 19, 15, 15, 16, 13, 17, 12, 18, 16, 17

a. Make a frequency table for these students listing scores from 12 to 20.
b. Find the median score.
c. Find the mode.
d. Find the mean.
 Measures of Central Tendency and Grouped Data 697

10. A questionnaire was distributed to 100 people. The table shows the time taken, in minutes,
to complete the questionnaire.
a. For this set of data, find: (1) the mean (2) the Interval Frequency
median (3) the mode
6 12
b. How are the three measures found in part a
5 20
related for these data?
4 36
3 20
2 12

11. A storeowner kept a tally of the sizes of suits purchased in the store, as shown
in the table.
a. For this set of data, find: Size of Suit Number Sold
(1) the total frequency (2) the mean (interval) (frequency)
(3) the median (4) the mode 48 1
b. Which measure of central tendency should the store- 46 1
owner use to describe the average suit sold? 44 3
42 5
40 3
38 8
36 2
34 2
12. Test scores for a class of 20 students are as follows:
93, 84, 97, 98, 100, 78, 86, 100, 85, 92, 72, 55, 91, 90, 75, 94, 83, 60, 81, 95
a. Organize the data in a table using 51–60 as the smallest interval.
b. Find the modal interval.
c. Find the interval that contains the median.
13. The following data consist of the weights, in pounds, of 35 adults:
176, 154, 161, 125, 138, 142, 108, 115, 187, 158, 168, 162
135, 120, 134, 190, 195, 117, 142, 133, 138, 151, 150, 168
172, 115, 148, 112, 123, 137, 186, 171, 166, 166, 179
a. Organize the data in a table, using 100–119 as the smallest interval.
b. Construct a frequency histogram based on the grouped data.
c. In what interval is the median for these grouped data?
d. What is the modal interval?
698 Statistics

16-6 QUARTILES, PERCENTILES, AND CUMULATIVE FREQUENCY

Quartiles
When the values in a set of data are listed in numerical order, the median sepa-
rates the values into two equal parts. The numbers that separate the set into four
equal parts are called quartiles.
To find the quartile values, we first divide the set of data into two equal parts
and then divide each of these parts into two equal parts.
The heights, in inches, of 20 students are shown in the following list. The
median, which is the average of the 10th and 11th data values, is shown here
enclosed in a box.
Lower half
______________________________ Upper half
_______________________________
| | | |

53, 60, 61, 63, 64, 65, 65, 65, 65, 66, 66, 67, 67, 68, 69, 70, 70, 71, 71, 73
↑
66
Median

Ten heights are listed in the lower half, 53–66. The middle value for these 10
heights is the average of the 5th and 6th values from the lower end, or 64.5. This
value separates the lower half into two equal parts.
Ten heights are also listed in the upper half, 66–73. The middle value for
these 10 heights is the average of the 5th and 6th values from the upper end, or
69.5. This value separates the upper half into two equal parts.
The 20 data values are now separated into four equal parts, or quarters.
_______________
| |
_______________
| |
_______________
| |
_______________
| |

53, 60, 61, 63, 64, 65, 65, 65, 65, 66, 66, 67, 67, 68, 69, 70, 70, 71, 71, 73
↑ ↑ ↑
64.5 66 69.5
Median
First quartile Second quartile Third quartile

The numbers that separate the data into four equal parts are the quartiles.
For this set of data:
1. Since one quarter of the heights are less than or equal to 64.5 inches, 64.5
is the lower quartile, or first quartile.
2. Since two quarters of the heights are less than or equal to 66 inches, 66 is
the second quartile. The second quartile is always the same as the median.
3. Since three quarters of the heights are less than or equal to 69.5 inches,
69.5 is the upper quartile, or third quartile.
Note: The quartiles are sometimes denoted Q1, Q2, and Q3.
 Quartiles, Percentiles, and Cumulative Frequency 699

Procedure
To find the quartile values for a set of data:
1. Arrange the data in ascending order from left to right.
2. Find the median for the set of data.The median is the second quartile value.
3. Find the middle value for the lower half of the data.This number is the first,
or lower, quartile value.
4. Find the middle value for the upper half of the data.This number is the
third, or upper, quartile value.

Note that when finding the first quartile, use all of the data values less than
or equal to the median, but do not include the median in the calculation.
Similarly, when finding the third quartile, use all of the data values greater than
or equal to the median, but do not include the median in the calculation.

Constructing a Box-and-Whisker Plot

A box-and-whisker plot is a diagram that uses the quartile values, together with
the maximum and minimum values, to display information about a set of data.
To draw a box-and-whisker plot, we use the following steps.
STEP 1. Draw a scale with numbers from the minimum to the maximum value
of a set of data. For example, for the set of heights of the 20 students,
the scale should include the numbers from 53 to 73.
STEP 2. Above the scale, place dots to represent the five numbers that are the
statistical summary for this set of data: the minimum value, the first
quartile, the median, the third quartile, and the maximum value.
For the heights of the 20 students, these numbers are 53, 64.5, 66, 69.5
and 73.

50 55 60 65 70 75

STEP 3. Draw a box between the dots that represent the lower and upper quar-
tiles, and a vertical line in the box through the point that represents the
median.

50 55 60 65 70 75
700 Statistics

STEP 4. Add the whiskers by drawing a line segment joining the dots that rep-
resent the minimum data value and the lower quartile, and a second
line segment joining the dots that represent the maximum data value
and the upper quartile.

50 55 60 65 70 75

The box indicates the ranges of the middle half of the set of data. The long
whisker at the left shows us that the data are more scattered at the lower than
at the higher end.

A graphing calculator can display a box-

and-whisker plot. Enter the data in L1, then go Plot1 Plot2 Plot3
to the STAT PLOT menu to select the type of On = Off
graph to draw. Ty p e :

ENTER: 2nd STAT PLOT 1 ENTER Xlist: L1

Freq: 1
 佡 佡 佡 佡 ENTER 

2nd L1  ALPHA 1

Now display the box-and-whisker plot by P1:L1

entering ZOOM 9 .

We can press TRACE and the right and

left arrow keys to display the minimum value,
first quartile, median, third quartile, and maxi-
mum value. Med=66

The five statistical summary can also be 1–Var Stats

<

I
n
displayed in 1-Var Stats. Scroll down to the last MinX=53
five values. Q 1= 6 4 . 5
Med=66
ENTER: STAT 佡 ENTER ENTER Q 3= 6 9 . 5
MaxX=73

EXAMPLE 1

Find the five statistical summary for the following set of data:
8, 5, 12, 9, 6, 2, 14, 7, 10, 17, 11, 8, 14, 5
 Quartiles, Percentiles, and Cumulative Frequency 701

Solution (1) Arrange the data in numerical order:

2, 5, 5, 6, 7, 8, 8, 9,10, 11, 12, 14, 14, 17
We can see that 2 is the minimum value and 17 is the maximum value.
(2) Find the median. Since there are 14 data values in the set, the median is
the average of the 7th and 8th values.
Median  8 1 9
2  8.5
Therefore, 8.5 is the second quartile.
(3) Find the first quartile. There are seven values less than 8.5. The middle
value is the 4th value from the lower end of the set of data, 6. Therefore, 6
is the first, or lower, quartile.
(4) Find the third quartile. There are seven values greater than 8.5. The middle
value is the 4th value from the upper end of the set of data, 12. Therefore,
12 is the third, or upper, quartile.

Answer The minimum is 2, first quartile is 6, the second quartile is 8.5, the third
quartile is 12, and the maximum is 17.

Note: The quartiles 6, 8.5, and 12 separate the data values into four equal parts
even though the original number of data values, 14, is not divisible by 4:
_____
| |
_____ ________
| | |
_________
| | |

2, 5, 5, 6, 7, 8, 8, 9, 10, 11, 12, 14, 14, 17

↑
8.5

The first and third quartile values, 6 and 12, are data values. If we think of each
of these as a half data value in the groups that they separate, each group con-
tains 312 data values, which is 25% of the total.

Percentiles
A percentile is a number that tells us what percent of the total number of data
values lies at or below a given measure.
Let us consider again the set of data values representing the heights of 20
students. What is the percentile rank of 65? To find out, we separate the data
into the values that are less than or equal to 65 and those that are greater than
or equal to 65, so that the four 65’s in the set are divided equally between the
two groups:
53, 60, 61, 63, 64, 65, 65, 65, 65, 66, 66, 67, 67, 68, 69, 70, 70, 71, 71, 73

Half of 4, or 2, of the 65’s are in the lower group and half are in the upper group.
702 Statistics

Since there are seven data values in the lower group, we find what percent
7 is of 20, the total number of values:
7
20 5 0.35 5 35%
Therefore, 65 is at the 35th percentile.
To find the percentile rank of 69, we separate the data into the values that
are less than or equal to 69 and those that are greater or equal to 69:
53, 60, 61, 63, 64, 65, 65, 65, 65, 66, 66, 67, 67, 68, 69, 70, 70, 71, 71, 73
Because 69 occurs only once, we will include it as half of a data value in the
lower group and half of a data value in the upper group. Therefore, there are 1412
or 14.5 data values in the lower group.
14.5
20 5 0.725 5 72.5%
Because percentiles are usually not written using fractions, we say that 69 is
at the 73rd percentile.

EXAMPLE 2

Find the percentile rank of 87 in the following set of 30 marks:

56, 65, 65, 67, 72, 73, 75, 77, 77, 78, 78, 78, 80, 80, 80,
82, 83, 85, 85, 85, 86, 87, 87, 87, 88, 90, 92, 93, 95, 98

Solution (1) Find the sum of the number of marks less than 87 and half of the number
of 87’s:
Number of marks less than 87  21
Half of the number of 87’s (0.5  3)  1.5
22.5
(2) Divide the sum by the total number of marks:
22.5
30 5 0.75
(3) Change the decimal value to a percent: 0.75  75%.

Answer: A mark of 87 is at the 75th percentile.

Note: 87 is also the upper quartile mark.

Cumulative Frequency
In a school, a final examination was given to all 240 students taking biology. The
test grades of these students were then grouped into a table. At the same time,
a histogram of the results was constructed, as shown below.
 Quartiles, Percentiles, and Cumulative Frequency 703

Interval Frequency HISTOGRAM

80
(test scores) (number)
60

Frequency
91–100 45
81–90 60 40
71–80 75 20
61–70 40
0
51–60 20 51–60 61–70 71–80 81–90 91–100
Test scores

From the table and the histogram, we can see that 20 students scored in the
interval 51–60, 40 students scored in the interval 61–70, and so forth. We can use
these data to construct a new type of histogram that will answer the question,
“How many students scored below a certain grade?”
By answering the following questions, we will gather some information
before constructing the new histogram:
1. How many students scored 60 or less on the test?
From the lowest interval, 51–60, we know that 20 students scored 60 or
less.
2. How many students scored 70 or less on the test?
By adding the frequencies for the two lowest intervals, 51–60 and 61–70,
we see that 20  40, or 60, students scored 70 or less.
3. How many students scored 80 or less on the test?
By adding the frequencies for the three lowest intervals, 51–60, 61–70, and
71–80, we see that 20  40  75, or 135, students scored 80 or less.
4. How many students scored 90 or less on the test?
Here, we add the frequencies in the four lowest intervals. Thus, 20  40 
75  60, or 195, students scored 90 or less.
5. How many students scored 100 or less on the test?
By adding the five lowest frequencies, 20  40  75  60  45, we see
that 240 students scored 100 or less. This result makes sense because 240
students took the test and all of them scored 100 or less.

Constructing a Cumulative Frequency Histogram

The answers to the five questions we have just asked were found by adding, or
accumulating, the frequencies for the intervals in the grouped data to find the
cumulative frequency. The accumulation of data starts with the lowest interval
of data values, in this case, the lowest test scores. The histogram that displays
these accumulated figures is called a cumulative frequency histogram.
704 Statistics

Interval Frequency Cumulative CUMULATIVE FREQUENCY

(test scores) (number) Frequency HISTOGRAM
240
91–100 45 240
210
81–90 60 195
71–80 75 135 180

Cumulative frequency
61–70 40 60
51–60 20 20 150

120

90

60

30

51 0
00
–6

–7

–8

–9
–1
51

51

51

51
Test scores
To find the cumulative frequency for each
CUMULATIVE FREQUENCY interval, we add the frequency for that inter-
HISTOGRAM val to the frequencies for the intervals with
100% 240
lower values. To draw a cumulative frequency
histogram, we use the cumulative frequencies
210
to determine the heights of the bars.
For our example of the 240 biology stu-
75% 180 dents and their scores, the frequency scale for
Cumulative frequency

the cumulative frequency histogram goes

Percent scale

150 from 0 to 240 (the total frequency for all of

the data). We can replace the scale of the
50% 120 cumulative frequency histogram shown above
with a different one that expresses the cumu-
90 lative frequency in percents. Since 240 stu-
dents represent 100% of the students taking
25% 60 the biology test, we write 100% to correspond
to a cumulative frequency of 240. Similarly,
30 since 0 students represent 0% of the students
taking the biology test, we write 0% to corre-
0% 0 spond to a cumulative frequency of 0.
If we divide the percent scale into four
0

51 0
00
–6

–7

–8

–9

equal parts, we can label the three added divi-

–1
51

51

51

51

sions as 25%, 50%, and 75%.

Test scores
 Quartiles, Percentiles, and Cumulative Frequency 705

Thus the graph relates each cumulative frequency to a percent of the total
number of biology students. For example, 120 students (half of the total num-
ber) corresponds to 50%.
Let us use the percent scale to answer the question, “What percent of the
students scored 70 or below on the test?” The height of each bar represents both
the number of students and the percent of the students who had scores at or
below the largest number in the interval represented by that bar. Since 25%, or
a quarter, of the scores were 70 or below, we say that 70 is an approximate value
for the lower quartile, or the 25th percentile.

CUMULATIVE FREQUENCY
HISTOGRAM
100% 240

210

75% Cumulative frequency 180

150
56%
50% 120

90

25% 60

30
8%
0% 0
0

51 0
00
–6

–7

–8

–9
–1
51

51

51

51

Test scores

From the histogram, we can see that about 56% of the students had scores
at or below 80. Thus, the second quartile, the median, is in the 51–80 interval. For
these data, the upper quartile is in the 51–90 interval.
From the histogram, we can also conveniently read the approximate per-
centiles for the scores that are the end values of the intervals. For example, to
find the percentile for a score of 60, the right-end score of the first interval, we
draw a horizontal line segment from the height of the first interval to the per-
cent scale, as shown by the dashed line in the histogram above. The fact that the
horizontal line crosses the percent scale at about one-third the distance between
0% and 25% tells us that approximately 8% of the students scored 60 or below
60. Thus, the 8th percentile is a good estimate for a score of 60.
706 Statistics

EXAMPLE 3

A reporter for the local newspaper is preparing an article on the ice cream
stores in the area. She listed the following prices for a two-scoop cone at 15
stores.
$2.48, $2.57, $2.30, $2.79, $2.25, $3.00, $2.82, $2.75,
$2.55, $2.98, $2.53, $2.40, $2.80, $2.50, $2.65

a. List the data in a stem-and-leaf diagram.

b. Find the median.
c. Find the first and third quartiles.
d. Construct a box-and-whisker plot.
e. Draw a cumulative frequency histogram.
f. Find the percentile rank of a price of $2.75.

Solution a. The first two digits in each price will be the Stem Leaf
stem. The lowest price is $2.25 and the highest
price is $3.00. 3.0 0
2.9 8
b. Since there are 15 prices, the median is the 8th 2.8 02
from the top or from the bottom. The median 2.7 59
is $2.57. 2.6 5
c. The middle value of the set of numbers below 2.5 0357
the median is the first quartile. That price is 2.4 08
$2.48. 2.3 0
2.2 5
The middle value of the set of numbers above
the median is the third quartile. That price is Key: 2.9  8  $2.98
$2.80.
d. Use a scale from $2.25 to $3.00. Place dots at $2.48, $2.57, and $2.80 for the
first quartile, the median, and the third quartile. Draw the box around the
quartiles with a vertical line through the median. Add the whiskers.

2.25 2.50 2.75 3.00

 Quartiles, Percentiles, and Cumulative Frequency 707

Cumulative CUMULATIVE FREQUENCY HISTOGRAM

Interval Frequency Frequency
15
3.00–3.09 1 15

Cumulative Frequency
2.90–2.99 1 14 12

2.80–2.89 2 13 9
2.70–2.79 2 11
2.60–2.69 1 9 6

2.50–2.59 4 8 3
2.40–2.49 2 4
0
2.30–2.39 1 2

2. .29

2. .39

2. .49

2. .69

2. .79

2. .89

9
.5

.9

.0
2.20–2.29 1 1

–2

–2

–2

–2

–2

–2

–2

–2

–3
20

20

20

20

20

20

20

20

20
2.

2.

2.
Price of a Two-Scoop Cone

e. Make a cumulative frequency table and draw the histogram. Use 2.20–2.29
as the smallest interval.
f. There are 9 data values below $2.75. Add 12 for the data value $2.75.
9 1
–
Percentile rank: 152  0.63  63%

Answers a. Diagram b. median  $2.57 c. first quartile  $2.48; third quartile  $2.80
d. Diagram e. Diagram f. 63rd percentile
Note: A cumulative frequency histogram can be drawn on a calculator just like
a regular histogram. In list L2, where we previously entered the frequencies
for each individual interval, we now enter each cumulative frequency.

EXERCISES
Writing About Mathematics
1. a. Is it possible to determine the percentile rank of a given score if the set of scores is
arranged in a stem-and-leaf diagram? Explain why or why not.
b. Is it possible to determine the percentile rank of a given score if the set of scores is
shown on a cumulative frequency histogram? Explain why or why not.
2. A set of data consisting of 23 consecutive numbers is written in numerical order from left to
right.
a. The number that is the first quartile is in which position from the left?
b. The number that is the third quartile is in which position from the left?
708 Statistics

Developing Skills
In 3–6, for each set of data: a. Find the five numbers of the statistical summary b. Draw a box-and-
whisker plot.
3. 12, 17, 20, 21, 25, 27, 29, 30, 32, 33, 33, 37, 40, 42, 44
4. 67, 70, 72, 77, 78, 78, 80, 84, 86, 88, 90, 92
5. 0, 0, 1, 1, 1, 1, 2, 2, 2, 2, 3, 3, 3, 3, 5, 7, 9, 9
6. 3.6, 4.0, 4.2, 4.3, 4.5, 4.8, 4.9, 5.0

In 7–9, data are grouped into tables. For each set of data:
a. Construct a cumulative frequency histogram.
b. Find the interval in which the lower quartile lies.
c. Find the interval in which the median lies.
d. Find the interval in which the upper quartile lies.
7. Interval Frequency 8. Interval Frequency 9. Interval Frequency

41–50 8 25–29 3 1–4 4

31–40 5 20–24 1 5–8 3
21–30 2 15–19 3 9–12 7
11–20 5 10–14 9 13–16 2
1–10 4 5–9 9 17–20 2

10. For the data given in the table:

Interval Frequency
a. Construct a cumulative frequency his-
togram. 21–25 5
b. In what interval is the median? 16–20 4
c. The value 10 occurs twice in the data. 11–15 6
What is the percentile rank of 10? 6–10 3
1–5 2

11. For the data given in the table: Interval Frequency

a. Construct a cumulative frequency his-
togram. 33–37 4
b. In what interval is the median? 28–32 3
c. In what interval is the upper quartile? 23–27 7
d. What percent of scores are 17 or less? 18–22 12
e. In what interval is the 25th percentile? 13–17 8
8–12 5
3–7 1
 Quartiles, Percentiles, and Cumulative Frequency 709

Applying Skills
12. A group of 400 students were asked CUMULATIVE FREQUENCY HISTOGRAM
to state the number of minutes that 100%
each spends watching television in 1 90%
day. The cumulative frequency his-
togram shown below summarizes 80%
the responses as percents.
70%
a. What percent of the students
questioned watch television for 60%
90 minutes or less each day?
50%
b. How many of the students watch
television for 90 minutes or less 40%
each day?
30%
c. In what interval is the upper
quartile? 20%
d. In what interval is the lower quar-
10%
tile?
e. If one of these students is picked

30

60

90

0
at random, what is the probability

12

15

18
1–

1–

1–

1–

1–

1–
that he or she watches 30 minutes
or less of television each day? Number of minutes
13. A journalism student was doing a study of the readability of the daily newspaper. She chose
several paragraphs at random and listed the number of letters in each of 88 words. She pre-
pared the following chart.
a. Copy the chart, adding a column that
Number of letters Frequency
lists the cumulative frequency
b. Find the median. 1 4
c. Find the first and third quartiles. 2 14
d. Construct a box-and whisker plot. 3 20
e. Draw a cumulative frequency his- 4 20
togram. 5 3
f. Find the percentile rank of a word with 6 18
7 letters.
7 5
8 2
9 1
10 1

14. Cecilia’s average for 4 years is 86. Her average is the upper quartile for her class of 250 stu-
dents. At most, how many students in her class have averages that are less than Cecilia’s?
710 Statistics

15. In the table at the right, data are given for the
Height Cumulative
heights, in inches, of 22 football players.
(inches) Frequency Frequency
a. Copy and complete the table.
77 2
b. Draw a cumulative frequency histogram.
76 2
c. Find the height that is the lower quartile.
75 7
d. Find the height that is the upper quartile.
74 5
73 3
72 2
71 1

16. The lower quartile for a set of data was 40. These data consisted of the heights, in inches, of
680 children. At most, how many of these children measured more than 40 inches?

In 17 and 18, select, in each case, the numeral preceding the correct answer.
17. On a standardized test, Sally scored at the 80th percentile. This means that
(1) Sally answered 80 questions correctly.
(2) Sally answered 80% of the questions correctly.
(3) Of the students who took the test, about 80% had the same score as Sally.
(4) Of the students who took the test, at least 80% had scores that were less than or
equal to Sally’s score.
18. For a set of data consisting of test scores, the 50th percentile is 87. Which of the following
could be false?
(1) 50% of the scores are 87. (3) Half of the scores are at least 87.
(2) 50% of the scores are 87 or less. (4) The median is 87.

16-7 BIVARIATE STATISTICS

We have been studying univariate statistics or statistics that involve a single set
of numbers. Statistics are often used to study the relationship between two dif-
ferent sets of values. For example, a dietician may want to study the relationship
between the number of calories from fat in a person’s diet and the level of cho-
lesterol in that person’s blood, or a merchant may want to study the relationship
between the amount spent on advertising and gross sales. Although these exam-
ples involving two-valued statistics or bivariate statistics require complex statis-
tical methods, we can investigate some of the properties of similar but simpler
problems by looking at graphs and by using a graphing calculator. A graph that
shows the pairs of values in the data as points in the plane is called a scatter plot.
 Bivariate Statistics 711

Correlation
We will consider five cases of two-valued statistics to investigate the relation-
ship or correlation between the variables based on their scatter plots.

CASE 1 The data has positive linear correlation. The points in the scatter plot
approximate a straight line that has a positive slope.

A driver recorded the number of gallons of gasoline used and the number
of miles driven each time she filled the tank. In this example, there is both cor-
relation and causation since the increase in the number of miles driven causes
the number of gallons of gasoline needed to increase.

Gallons 7.2 5.8 7.0 5.5 5.6 7.1 6.0 4.4 5.0 6.2 4.7 5.7
Miles 240 188 226 193 187 235 202 145 167 212 154 188

This scatter plot can be dupli- 250

cated on your graphing calculator.
Enter the number of miles as L1 and 225

Miles
the number of gallons of gasoline as 200
corresponding entries in L2. The miles 175
will be graphed as x-values and the
150
gallons of gasoline as y-values. First,
turn on Plot 1:
3 4 5 6 7 8 9
Gallons of Gasoline

ENTER: 2nd STAT PLOT 1 ENTER  ENTER  2nd L1

 2nd L2

DISPLAY:
Plot1 Plot2 Plot3
On = Off
Ty p e :

Xlist: L1
Ylist: L2
Mark: + .
712 Statistics

Now use ZoomStat from the ZOOM menu to construct a window that will
include all values of x and y.

ENTER: ZOOM 9

DISPLAY:

CASE 2 The data has moderate positive correlation. The points in a scatter plot
do not lie in a straight line but there is a general tendency for the values of y to
increase as the values of x increase.
Last month, each student in an English class was required to choose a book
to read. The teacher recorded, for each student in the class, the number of days
spent reading the book and the number of pages in the book.

Days 8 14 12 26 9 17 28 13 15 30 18 20
Pages 225 300 298 356 200 412 205 215 310 357 209 250
Days 29 22 17 14 11 14 22 19 16 7 18 30
Pages 314 288 256 225 232 256 300 305 276 172 318 480

500 While the books with more pages

475 may have required more time, some stu-
450
425 dents read more rapidly and some spent
400 more time each day reading. The graph
375 shows that, in general, as the number of
350 days needed to read a book increased,
Pages

325 the number of pages that were read also

300 increased.
275
250
225
200
175

5 7 9 11 13 15 17 19 21 23 25 27 29 31

Days
 Bivariate Statistics 713

CASE 3 The data has no correlation.

Before giving a test, a teacher asked each student how many minutes each
had spent the night before preparing for the test. After correcting the test, she
prepared the table below which compares the number of minutes of study to the
number of correct answers.

Minutes
of Study 20 15 40 5 10 25 30 12 5 20 35 40

Correct
Answers 15 10 3 19 16 6 12 3 5 8 16 14

The graph shows that 20

there is no correlation 18
between the time spent

Correct Answers
16
studying just before the test 14
and the number of correct 12
answers on the test. 10
8
6
4
2
0
5 10 15 20 25 30 35 40 45
Minutes of Study

CASE 4 The data has moderate negative correlation. The points in a scatter plot
do not lie in a straight line but there is a general tendency for the values of y to
decrease as the values of x increase.

A group of children go to an after-school program at a local youth club. The

director of the program keeps a record, shown below, of the time, in minutes,
each student spends playing video games and doing homework.

Games 20 30 90 60 30 50 70 40 80 60
Homework 50 60 10 40 40 35 15 30 30 10

In this instance, the unit of measure, minutes, is found in the problem rather
than in the table. To create meaningful graphs, always include a unit of measure
on the horizontal and vertical axes.
714 Statistics

70

Minutes Doing Homework

60
50
40
30
20
10
0
0 10 20 30 40 50 60 70 80 90 100
Minutes Playing Games

The graph shows that, in general, as the number of minutes spent playing
video games increases, the number of minutes spent doing homework decreases.
CASE 5 The data has negative linear correlation. The points in the scatter plot
approximate a straight line that has a negative slope.
A long-distance truck driver travels 500 miles each day. As he passes
through different areas on his trip, his average speed and the length of time he
drives each day vary. The chart below shows a record of average speed and time
for a 10-day period.

Speed 50 64 68 60 54 66 70 62 64 58
Time 10 7.9 7.5 8.5 9.0 7.0 7.1 8.0 8.2 9.0

11 In this case, the increase in the average speed causes

10.5 the time required to drive a fixed distance to decrease.
10 This example indicates both negative correlation and
9.5 causation.
Time in Hours

9 It is important to note that correlation is not the same

8.5 as causation. Correlation is an indication of the strength
8 of the linear relationship or association between the vari-
7.5 ables, but it does not mean that changes in one variable
7 are the cause of changes in the other. For example, sup-
6.5 pose a study found there was a strong positive correlation
between the number of pages in the daily newspaper and
50 55 60 65 70 75 the number of voters who turn out for an election. One
would not be correct in concluding that a greater number
Speed in Miles per Hour of pages causes a greater turnout. Rather, it is likely that
the urgency of the issues is reflected in the increase of
both the size of the newspaper and the size of the turnout.
 Bivariate Statistics 715

Another example where there is no causation occurs in time series or data

that is collected at regular intervals over time. For instance, the population of
the U.S. recorded every ten years is an example of a time series. In this case, we
cannot say that time causes a change in the population. All we can do is note a
general trend, if any.

Line of Best Fit

When it makes sense to consider one variable as the independent variable and
the other as the dependent variable, and the data has a linear correlation (even
if it is only moderate correlation), the data can be represented by a line of best
fit. For example, we can write an equation for the data in Case 1. Enter the data
into L1 and L2 if it is not already there. Find the mean values for x, the number
of miles driven, and for y, the number of gallons of gasoline used. Then use
2-Var Stats from the STAT CALC menu:

ENTER: STAT 佡 2 ENTER

DISPLAY:
2–Var Stats
–
x=5.85
x=70.2
x2=419.88
Sx=.9150260801
x=.8760707734
In=12
<

The calculator gives x–  5.85 and, by pressing the down arrow key, y–  194.75.
We will use these mean values, (5.85, 194.75), as one of the points on our line.
We will choose one other data point, for example (7.1, 235), as a second point
and write the equation of the line using the slope-intercept form y  mx  b.
First find the slope:
y 2y 194.75 2 235 240.25
m  x22 2 x11 5 5.85 2 7.1 5 21.25 5 32.2
Now use one of the points to find the y-intercept:
194.75  32.2(5.85)  b
194.75  188.37  b
6.38  b
Round the values to three significant digits. A possible equation for a line of
best fit is y  32.2x  6.38.
The calculator can also be used to find a line called the regression line to fit
a bivariate set of data. Use the LinReg(ax+b) function in the STAT CALC
menu.
716 Statistics

ENTER: STAT 佡 4 ENTER

DISPLAY:
LinReg
y = a=x + b
a=32.67643865
b=3.592833876

If we round the values to three significant digits, the equation of the regres-
sion line is y  32.7x  3.59. In this case, the difference between these two equa-
tions is negligible. However, this is not always the case. The regression line is a
special line of best fit that minimizes the square of the vertical distances to each
data point.
We can compare these two equations with the actual data. Graph the scat-
ter plot of the data using ZoomStat. Then write the two equations in the Y=
menu.

ENTER: Y 32.2 X,T,,n  6.38 ENTER

32.7 X,T,,n  3.59 GRAPH

DISPLAY:

Notice that the lines are very close and do approximate the data.
Note 1: The equation of the line of best fit is very sensitive to rounding. Try to
round the coefficients of the line of best fit to at least three significant digits
or to whatever the test question asks.
Note 2: A line of best fit is appropriate only for data that exhibit a linear pattern.
In more advanced courses, you will learn how to deal with nonlinear patterns.
These equations can be used to predict values. For example, if the driver has
driven 250 miles before filling the tank, how many gallons of gasoline should be
needed? We will use the equation from the calculator.
y  32.7x  3.59
250  32.7x  3.59
246.41  32.7x
7.535474006  x
 Bivariate Statistics 717

It is reasonable to say that the driver can expect to need about 7.5 gallons of
gasoline.
What we just did is called extrapolation, that is, using the line of best fit
to make a prediction outside of the range of data values. Using the line of
best fit to make a prediction within the given range of data values is called
interpolation.
In general, interpolation is usually safe, while care should be taken when
extrapolating. The observed correlation pattern may not be valid outside of the
given range of data values. For example, consider the scatter plot of the popula-
tion of a town shown below. The population grew at a constant rate during the
years in which the data was gathered. However, we do not expect the popula-
tion to continue to grow forever, and thus, it may not be possible to extrapolate
far into the future.

280
270
260
Population (in thousands)

250
240
230
220
210
200
190
180
170
160

1960 1965 1970 1975 1980 1985 1990 1995 2000 2005

Year

Keep in Mind In general, when a given relationship involves two sets of data:
1. In some cases a straight line, a line of best fit, can be drawn to approxi-
mate the relationship between the data sets.
2. If a line of best fit has a positive slope, the data has positive linear correla-
tion.
3. If the line of best fit has a negative slope, the data has negative linear cor-
relation.
4. A line of best fit can be drawn through (x–, y–), the point whose coordinates
are the means of the given data. Any data point that appears to lie on
or near the line of best fit can be used as a second point to write the
equation.
718 Statistics

5. A calculator can be used to find the regression line as the line of best fit.
6. When the graphed data points are so scattered that it is not possible to
draw a straight line that approximates the given relationship, the data has
no correlation.
To study bivariate data without using a graphing calculator:
1. Make a table that lists the data.
2. Plot the data as points on a graph.
3. Find the mean of each set of data and locate the point (x–, y–) on the graph.
4. Draw a line that best approximates the data.
5. Choose the point (x–, y–) and one other point or any two points that are on
or close to the line that you drew. Use these points to write an equation of
the line.
6. Use the equation of the line to predict related outcomes.
To study bivariate data using a graphing calculator:
1. Enter the data into L1 and L2 or any two lists in the memory of the calcu-
lator.
2. Use STAT PLOT to turn on a plot and to choose the type of plot needed.
Enter the names of the lists in which the data is stored and choose the
mark to be used for each data point.
3. Use ZoomStat to choose a viewing window that shows all of the data
points.
4. Find the regression line using LinReg(ax+b) from the STAT CALC menu.
5. Enter the equation of the regression line in the Y= menu and use GRAPH
to show the relationship between the data and the regression line.
6. Use the equation of the line to predict related outcomes.
In this course, we have found a line of best fit by finding a line that seems to
represent the data or by using a calculator. In more advanced courses in statis-
tics, you will learn detailed methods for finding the line of best fit.

EXAMPLE 1

The table below shows the number of calories and the number of grams of car-
bohydrates in a half-cup serving of ten different canned or frozen vegetables.

Carbohydrates 9 23 4 5 19 8 12 7 13 17
Calories 45 100 20 25 110 35 50 30 70 80
 Bivariate Statistics 719

a. Draw a scatter plot on graph paper. Let the horizontal axis represent
grams of carbohydrates and the vertical axis represent the number of
calories.
b. Find the mean number of grams of carbohydrates in a serving of vegetables
and the mean number of calories in a serving of vegetables.
c. On the graph, draw a line that approximates the data in the table, and deter-
mine its equation.
d. Enter the data in L1 and L2 on your calculator and find the linear regression
equation, LinReg(ax+b).
e. Use each equation to find the expected number of calories in a serving of
vegetables with 20 grams of carbohydrates. Compare the answers.

Solution a. 120
110
100
90
Calories

80
70
60
50
40
30
20
0 5 10 15 20 25

Grams of Carbohydrates

b. Enter the number of grams of carbohydrates in L1 and the number of calo-

ries in L2. Find x– and y–, using 2-Var Stats.

ENTER: STAT 佡 2 ENTER

The means of the x- and y-coordinates are x–  11.7 and y–  56.5.

Locate the point (11.7, 56.5) on the graph.
720 Statistics

c. 120
110
100
90
80
70

Calories
60 (x, y)
50
40
30
20
0 5 10 15 20 25
Grams of Carbohydrates

The line we have drawn seems to go through the point (4, 20). We will use
this point and the point with the mean values, (11.7, 56.5), to write an equa-
tion of a line of best fit.
m 5 56.5 2 20 36.5
11.7 2 4 5 7.7 < 4.74 y 5 mx 1 b
20 5 36.5
7.7 (4) 1 b
1.05 < b
An equation of a best fit line is y  4.74x  1.05.
d. The data are in L1 and L2.

ENTER: STAT 佡 4 ENTER

DISPLAY:
LinReg
y = a=x + b
a=4.885506842
b=-.6604300475

The equation of the regression line is y = 4.89x  0.660.

e. Let x  20.
Use the equation from c. Use the equation from d.
y  4.74x  1.05 y = 4.89x  0.660
y  4.74(20)  1.05 y  4.89(20)  0.660
y  95.85 y  97.14
The two equations give very similar results. It would be reasonable to
say that we could expect the number of calories to be about 96 or close
to 100.
 Bivariate Statistics 721

EXERCISES
Writing About Mathematics
1. a. Give an example of a set of bivariate data that has negative correlation.
b. Do you think that the change in the independent variable in your example causes the
change in the dependent variable?
2. Explain the purpose of finding a line of best fit.

Applying Skills
3. When Gina bought a new car, she decided to keep a record of how much gas she uses. Each
time she puts gas in the car, she records the number of gallons of gas purchased and the
number of miles driven since the last fill-up. Her record for the first 2 months is as follows:

Gallons of gas 10 12 9 6 11 10 8 12 10 7
Miles driven 324 375 290 190 345 336 250 375 330 225

a. Draw a scatter plot of the data. Let the horizontal axis represent the number of gallons
of gas and the vertical axis represent the number of miles driven.
b. Does the data have positive, negative, or no correlation?
c. Is this a causal relationship?
d. Find the mean number of gallons of gasoline per fill-up.
e. Find the mean number of miles driven between fill-ups.
f. Locate the point that represents the mean number of gallons of gasoline and the mean
number of miles driven. Use (0, 0) as a second point. Draw a line through these two
points to approximate the data in the table.
g. Use the line drawn in part d to approximate the number of miles Gina could drive on
3 gallons of gasoline.
4. Gemma made a record of the cost and length of each of the 14 long-distance telephone calls
that she made in the past month. Her record is given below.

Minutes 3.7 1.0 19.6 0.8 4.3 34.8 2.9

Cost $0.35 $0.11 $2.12 $0.09 $0.47 $3.78 $0.24
Minutes 2.5 7.1 10.9 5.8 1.5 1.4 8.0
Cost $0.27 $0.79 $1.21 $0.65 $0.20 $0.17 $0.89

a. Draw a scatter plot of the data on graph paper. Let the horizontal axis represent the
number of minutes, and the vertical axis represent the cost of the call.
b. Does the data have positive, negative, or no correlation?
c. Is this a causal relationship?
722 Statistics

d. Find the mean number of minutes per call.

e. Find the mean cost of the calls.
f. On the graph, draw a line of best fit for the data in the table and write its equation.
g. Use a calculator to find the equation of the regression line.
h. Approximate the cost of a call that lasted 14 minutes using the equation written in d.
i. Approximate the cost of a call that lasted 14 minutes using the equation written in e.
5. A local store did a study comparing the cost of a head of lettuce with the number of heads
sold in one day. Each week, for five weeks, the price was changed and the average number
of heads of lettuce sold per day was recorded. The data is shown in the chart below.

Cost per Head of Lettuce $1.50 $1.25 $0.90 $1.75 $0.50

No. of Heads Sold 48 52 70 42 88

a. Draw a scatter plot of the data. Let the horizontal axis represent the cost of a head of
lettuce and the vertical axis represent the number of heads sold.
b. Does the data have positive, negative, or no correlation?
c. Is this a causal relationship?
d. Find the mean cost per head.
e. Find the mean number of heads sold.
f. On the graph, draw a line that approximates the data in the table.
g. What appears to be the result of raising the price of a head of lettuce?
6. The chart below shows the recorded heights in inches and weights in pounds for the last 24
persons who enrolled in a health club.

Height Weight Height Weight Height Weight Height Weight

69 160 75 180 66 145 71 165

67 160 76 155 66 130 66 155
63 135 70 175 68 160 67 140
73 185 73 170 68 140 78 210
71 215 68 190 72 170 72 160
79 225 74 190 77 195 69 145

a. Draw a scatter plot on graph paper to display the data.

b. Does the data have positive, negative, or no linear correlation?
c. Is this a causal relationship?
d. Draw and find the equation of a line of best fit. Use (77, 195) as a second point.
e. Use a calculator to find the linear regression line.
 Bivariate Statistics 723

f. According to the equation written in d., if the next person who enrolls in the health
club is 62 inches tall, what would be the expected weight of that person?
g. According to the equation written in e., if the next person who enrolls in the health
club weighs 200 pounds, what would be the expected height of that person?
7. The chart below shows the number of millions of cellular telephones in use in the United
States by year from 1994 to 2003.

Year ’94 ’95 ’96 ’97 ’98 ’99 ’00 ’01 ’02 ’03
Phones 24.1 33.8 44.0 55.3 69.2 86.0 109.5 128.3 140.8 158.7

Let L1 be the number of years after 1990: 4, 5, 6, 7, 8, 9, 10, 11, 12, 13.
a. Draw a scatter plot on graph paper to display the data.
b. Does the data have positive, negative, or no linear correlation?
c. Is this a causal relationship?
d. Draw and find the equation of a line of best fit.
e. On the graph, draw a line that approximates the data in the table, and determine its
equation. Use (6, 44.0) as a second point.
f. If the line of best fit is approximately correct for years beyond 2003, estimate how
many cellular phones will be in use in 2007.
8. The chart below shows, for the last 20 Supreme Court Justices to have left the court before
2000, the age at which the judge was nominated and the number of years as a Supreme
Court judge.

Age 47 64 62 62 59 55 54 45 43 56
Years 15 16 24 17 24 4 3 31 23 5
Age 50 56 62 59 50 56 57 49 49 62
Years 33 16 16 7 18 7 13 6 12 1

a. Draw a scatter plot on graph paper to display the data.

b. Does the data have positive, negative, or no linear correlation?
c. Is there a causal relationship?
d. Draw and find an equation of a line of best fit.
e. Use a calculator to find the linear regression line.
f. Do you think that the data, the line of best fit, the regression line, or none of these
could be used to approximate the number of years as a Supreme Court justice for the
next person to retire from that office?
724 Statistics

9. A cook was trying different recipes for potato salad and comparing the amount of dressing
with the number of potatoes given in the recipe. The following data was recorded.

Number of Potatoes 7 4 2 8 6 7 5 4
7 3
Cups of Dressing 112 8 4 114 1 134 118 3
4

a. Draw a scatter plot on graph paper to display the data.

b. Does the data have positive, negative, or no linear correlation?
c. Draw and find the equation of a line of best fit. Use A 4, 78 B as a second point.
d. Use a calculator to find the linear regression line.
e. According to the equation written in c., if the cook needs to use 10 potatoes to have
enough salad, approximately how many cups of dressing are needed?

CHAPTER SUMMARY
Statistics is the study of numerical data. In a statistical study, data are col-
lected, organized into tables and graphs, and analyzed to draw conclusions.
Data can either be quantitative or qualitative. Quantitative data represents
counts or measurements. Qualitative data represents categories or qualities.
In an experiment, a researcher imposes a treatment on one or more groups.
The treatment group receives the treatment, while the control group does not.
Tables and stem-and-leaf diagrams are used to organize data. A table
should have between five and fifteen intervals that include all data values, are
of equal size, and do not overlap.
A histogram is a bar graph in which the height of a bar represents the fre-
quency of the data values represented by that bar.
A cumulative frequency histogram is a bar graph in which the height of the
bar represents the total frequency of the data values that are less than or equal
to the upper endpoint of that bar.
The mean, median, and mode are three measures of central tendency. The
mean is the sum of the data values divided by the total frequency. The median
is the middle value when the data values are placed in numerical order. The
mode is the data value that has the largest frequency.
Quartile values separate the data into four equal parts. A box-and-whisker
plot displays a set of data values using the minimum, the first quartile, the
median, the third quartile, and the maximum as significant measures. The per-
centile rank tells what percent of the data values lie at or below a given mea-
sure.
In two-valued statistics or bivariate statistics, a relation between two differ-
ent sets of data is studied. The data can be graphed on a scatter plot. The data
may have positive, negative, or no correlation. Data that has positive or negative
linear correlation can be represented by a line of best fit.
 Review Exercises 725

The line of best fit can be used to predict values not in the included data set.
Interpolation is predicting within the given data range. Extrapolation is pre-
dicting outside of the given data range.

VOCABULARY

16-1 Data • Statistics • Descriptive statistics • Qualitative data •

Quantitative data • Census • Sample • Bias • Experiment • Treatment
group • Control group • Placebo effect • Placebo • Blinding • Single-
blind experiment • Double-blind experiment
16-2 Tally • Frequency • Total frequency • Frequency distribution table •
Group • Interval • Grouped data • Range • Stem-and-leaf diagram •
16-3 Histogram • Frequency histogram
16-4 Average • Measures of central tendency • Mean • Arithmetic mean •
Numerical average • Median • Mode • Bimodal • Linear
transformation of a data set
16-5 Group mode • Modal interval
16-6 Quartile • Lower quartile • First quartile • Second quartile • Upper
quartile • Third quartile • Box-and-whisker plot • Five statistical
summary • Percentile • Cumulative frequency • Cumulative frequency
histogram
16-7 Univariate statistics • Two-valued statistics • Bivariate statistics •
Scatter plot • Correlation • Causation • Time series • Line of best fit •
Regression line • Extrapolation • Interpolation

REVIEW EXERCISES

1. Courtney said that the mean of a set of consecutive integers is the same as
the median and that the mean can be found by adding the smallest and
the largest numbers and dividing the sum by 2. Do you agree with
Courtney? Explain why or why not.
2. A set of data contains N numbers arranged in numerical order.
a. When is the median one of the numbers in the set of data?
b. When is the median not one of the numbers in the set of data?
3. For each of the following sets of data, find: a. the mean b. the median
c. the mode (if one exists)
(1) 3, 4, 3, 4, 3, 5 (2) 1, 3, 5, 7, 1, 2, 4
(3) 9, 3, 2, 8, 3, 3 (4) 9, 3, 2, 3, 8, 2, 7
4. Express, in terms of y, the mean of 3y  2 and 7y  18.
726 Statistics

5. For the following data:

78, 91, 60, 65, 81, 72, 78, 80, 65, 63, 59, 78, 78, 54, 87, 75, 77
a. Use a stem-and-leaf diagram to organize the data.
b. Draw a histogram, using 50–59 as the lowest interval.
c. Draw a cumulative frequency histogram.
d. Draw a box-and-whisker plot.
6. The weights, in kilograms, of five adults are 53, 72, 68, 70, and 72.
a. Find: (1) the mean (2) the median (3) the mode
b. If each of the adults lost 5 kilograms, find, for the new set of weights:
(1) the mean (2) the median (3) the mode
7. Steve’s test scores are 82, 94, and 91. What grade must Steve earn on a
fourth test so that the mean of his four scores will be exactly 90?
8. From Monday to Saturday of a week in May, the recorded high tempera-
ture readings were 72°, 75°, 79°, 83°, 83°, and 88°. For these data, find:
a. the mean b. the median c. the mode
9. Paul worked the following numbers of hours each week over a 20-week
period:
15, 3, 7, 6, 2, 14, 9, 25, 8, 12, 8, 8, 15, 0, 8, 12, 28, 10, 14, 10
a. Organize the data in a frequency table, using 0–5 as the lowest interval.
b. Draw a frequency histogram of the data.
c. In what interval does the median lie?
d. Which interval contains the lower quartile?
10. The table shows the scores of 25
test papers. Cumulative
Score Frequency Frequency
a. Is the data univariate or
bivariate? 60 1
b. Find the mean score. 70 9
c. Find the median score. 80 8
d. Find the mode. 90 2
e. Copy and complete the table. 100 5
f. Draw a cumulative frequency
histogram.
g. Find the percentile rank of 90.
h. What is the probability that a paper chosen at random has a score of
80?
 Review Exercises 727

11. The electoral votes cast for the winning presidential candidate in elections
from 1900 to 2004 are as follows:
292, 336, 321, 435, 277, 404, 382, 444, 472, 523, 449, 432, 303, 442,
457, 303, 486, 301, 520, 297, 489, 525, 426, 370, 379, 271, 286
a. Organize the data in a stem-and-leaf diagram. (Use the first digit as the
stems, and the last two digits as the leaves.)
b. Find the median number of electoral votes cast for the winning candi-
date.
c. Find the first-quartile and third-quartile values.
d. Draw a box-and-whisker plot to display the data.
12. The ages of 21 high school students are shown in
Age Frequency
the table at the right.
a. What is the median age? 18 1
b. What is the percentile rank of age 15? 17 4
c. When the ages of these 21 students are com- 16 2
bined with the ages of 20 additional students, 15 7
the median age remains unchanged. What is
14 2
the smallest possible number of students under
16 in the second group? 13 5

13. For each variable, determine if it is qualitative or

quantitative.
a. Major in college b. GPA in college
c. Wind speed of a hurricane d. Temperature of a rodent
e. Yearly profit of a corporation f. Number of students late to class
g. Zip code h. Employment status
14. Researchers looked into a possible relationship between alcoholism and
pneumonia. They conducted a study of 100 current alcoholics, 50 former
alcoholics, and 1,000 non-alcoholics who were hospitalized for a mild form
of pneumonia. The researchers found that 30% of alcoholics and 30% of
former alcoholics, versus only 15% of the non-alcoholics developed a
more dangerous form of pneumonia. The researchers concluded that alco-
holism raises the risk for developing pneumonia.
Discuss possible problems with this study.
728 Statistics

15. Aurora buys oranges every week. The accompanying table lists the weights
and the costs of her last 10 purchases of oranges.

Weight (lb) 2.2 1.2 3.6 4.5 1.0 2.5 1.8 5.0 3.5 1.7
Cost ($) 1.22 0.60 1.04 1.58 0.50 0.89 0.95 1.88 1.46 0.70

a. Is the data univariate or bivariate?

b. Draw a scatter plot of the data on graph paper. Let the horizontal axis
represent the weights of the oranges and the vertical axis the costs.
c. Is there a correlation between the weight and the cost of the oranges?
If so, is it positive or negative?
d. If the price is determined by the number of oranges purchased, do the
variables have a causal relationship? Explain your answer.
e. On the graph, draw a line of best fit that approximates the data in the
table and write its equation.
f. Use the equation written in d to approximate the cost of 4 pounds of
oranges.
16. Explain why the graph on NET INCOME OF
the right is misleading. XYZ COMPANY
(Hint: In accounting, num- (in thousands) $12,000
bers enclosed by parenthe-
ses denote negative
$8,050
numbers.) $7,100
$5,123
$(4,000)

2002 2003 2004 2005 2006

Exploration
a. Marny took the SAT in 2004 and scored a 1370. She was in the 94th
percentile. Jordan took the SAT in 2000 and scored 1370. He was in the
95th percentile. Explain how this is possible.
b. Taylor’s class rank stayed the same even though he had a cumulative
grade point average of 3.4 one semester and 3.8 the next semester.
Explain how this is possible.
 Cumulative Review 729

CUMULATIVE REVIEW CHAPTERS 1–16

Part I
Answer all questions in this part. Each correct answer will receive 2 credits. No
partial credit will be allowed.
1. When the domain is the set of integers, the solution set of the inequality
0 , 0.1x 2 0.4 # 0.2 is
(1) { } (2) {4, 5} (3) {4, 5, 6} (4) {5, 6}
2. The product (2a  3)(2a  3) can be written as
(1) 2a2  9 (3) 4a2  9
(2) 4a  9
2
(4) 4a2  12a  9
3. When 0.00034 is written in the form 3.4  10n, the value of n is
(1) 3 (2) 4 (3) 3 (4) 4
4. When x3 1 12 5 x 2 2
6 , x equals
(1) 5 (2) 1 (3) 12 (4) 1
5. The mean of the set of even integers from 2 to 100 is
(1) 49 (2) 50 (3) 51 (4) 52
6. The probability that 9 is the sum of the numbers that appear when two
dice are rolled is
(1) 64 2
(2) 36 (3) 62 4
(4) 36
7. If the circumference of a circle is 12 centimeters, then the area of the circle
is
(1) 36 square centimeters (3) 36
p square centimeters
(2) 144 square centimeters (4) 144
p square centimeters
8. Which of the following is not an equation of a function?
(1) y  3x  2 (3) y2  x
(2) y  x  3x  1
2
(4) y  x
9. The value of 10P8 is
(1) 80 (2) 90 (3) 1,814,400 (4) 3,628,800
10. Which of the following is an equation of a line parallel to the line whose
equation is y 5 22x 1 4?
(1) 2x  y  7 (2) y  2x  7 (3) 2x  y  7 (4) y  2x  7

Part II
Answer all questions in this part. Each correct answer will receive 2 credits.
Clearly indicate the necessary steps, including appropriate formula substitu-
730 Statistics

tions, diagrams, graphs, charts, etc. For all questions in this part, a correct numer-
ical answer with no work shown will receive only 1 credit.
11. In a bridge club, there are three more women than men. How many per-
sons are members of the club if the probability that a member chosen at
random, is a woman is 35?
12. Find to the nearest degree the measure of the smallest angle in a right tri-
angle whose sides measure 12, 35, and 37 inches.

Part III
Answer all questions in this part. Each correct answer will receive 3 credits.
Clearly indicate the necessary steps, including appropriate formula substitu-
tions, diagrams, graphs, charts, etc. For all questions in this part, a correct numer-
ical answer with no work shown will receive only 1 credit.
13. The lengths of the sides of a triangle are in the ratio 3 : 5 : 6. The perimeter
of the triangle is 49.0 meters. What is the length of each side of the triangle?
14. Huy worked on an assignment for four days. Each day he worked half as
long as he worked the day before and spent a total of 3.75 hours on the
assignment.
a. How long did Huy work on the assignment each day?
b. Find the mean number of hours that Huy worked each day.

Part IV
Answer all questions in this part. Each correct answer will receive 4 credits.
Clearly indicate the necessary steps, including appropriate formula substitu-
tions, diagrams, graphs, charts, etc. For all questions in this part, a correct numer-
ical answer with no work shown will receive only 1 credit.
15. The perimeter of a garden is 16 feet. Let x represent the width of the garden.
a. Write an equation for the area of the land, y, in terms of x.
b. Sketch the graph of the equation that you wrote in a.
c. What is the maximum area of the land?
16. Each morning, Malcolm leaves for school at 8:00 o’clock. His brother
Marvin leaves for the same school at 8:15. Malcolm walks at 2 miles an
hour and Marvin rides his bicycle at 8 miles an hour. They follow the same
route to school and arrive at the same time.
a. At what time do Malcolm and Marvin arrive at school?
b. How far is the school from their home?