bar67388_ch08_746-760 12/02/06 19:52 Page 760

760 CHAPTER 8 SEQUENCES, INDUCTION, AND PROBABILITY

53. Eight distinct points are selected on the circumference of a (C) The distinct positions are not taken into consideration,
circle. but either Mike or Ken, but not both, must start?
(A) How many chords can be drawn by joining the points in
all possible ways?
(B) How many triangles can be drawn using these eight
points as vertices?
(C) How many quadrilaterals can be drawn using these
eight points as vertices?
54. Five distinct points are selected on the circumference of a
circle.
(A) How many chords can be drawn by joining the points in
all possible ways?
(B) How many triangles can be drawn using these five
58. How many committees of four people are possible from a
points as vertices?
group of nine people if
55. How many ways can two people be seated in a row of five (A) There are no restrictions?
chairs? Three people? Four people? Five people? (B) Both Juan and Mary must be on the committee?
(C) Either Juan or Mary, but not both, must be on the com-
56. Each of two countries sends five delegates to a negotiating
mittee?
conference. A rectangular table is used with five chairs
on each long side. If each country is assigned a long side 59. A 5-card hand is dealt from a standard 52-card deck. Which
of the table, how many seating arrangements are possible? is more likely: the hand contains exactly one king or the
[Hint: Operation 1 is assigning a long side of the table to hand contains no hearts?
each country.]
60. A 10-card hand is dealt from a standard 52-card deck.
57. A basketball team has five distinct positions. Out of eight Which is more likely: all cards in the hand are red or the
players, how many starting teams are possible if hand contains all four aces?
(A) The distinct positions are taken into consideration?
(B) The distinct positions are not taken into consideration?

8-5 Sample Spaces and Probability

Z Experiments
Z Sample Spaces and Events
Z Probability of an Event
Z Equally Likely Assumption
Z Empirical Probability

This section provides an introduction to probability, a topic to which whole books and
courses are devoted. Probability involves many subtle notions, and care must be taken
at the beginning to understand the fundamental concepts on which the subject is based.
First, we develop a mathematical model for probability studies. Our development,
because of space, must be somewhat informal. More formal and precise treatments
can be found in books on probability.
bar67388_ch08_761-779 12/02/06 19:54 Page 761

S E C T I O N 8–5 Sample Spaces and Probability 761

Z Experiments
Our first step in constructing a mathematical model for probability studies is to
describe the type of experiments on which probability studies are based. Some types
of experiments do not yield the same results, no matter how carefully they are repeated
under the same conditions. These experiments are called random experiments. Famil-
iar examples of random experiments are flipping coins, rolling dice, observing the fre-
quency of defective items from an assembly line, or observing the frequency of deaths
in a certain age group.
Probability theory is a branch of mathematics that has been developed to deal
with outcomes of random experiments, both real and conceptual. In the work that fol-
lows, the word experiment will be used to mean a random experiment.

Z Sample Spaces and Events

Associated with outcomes of experiments are sample spaces and events. Our second
step in constructing a mathematical model for probability studies is to define these
two terms. Set concepts will be useful in this regard.
Consider the experiment, “A single six-sided die is rolled.” What outcomes might
we observe? We might be interested in the number of dots facing up, or whether the
number of dots facing up is an even number, or whether the number of dots facing
up is divisible by 3, and so on. The list of possible outcomes appears endless. In gen-
eral, there is no unique method of analyzing all possible outcomes of an experiment.
Therefore, before conducting an experiment, it is important to decide just what out-
comes are of interest.
In the die experiment, suppose we limit our interest to the number of dots fac-
ing up when the die comes to rest. Having decided what to observe, we make a list
of outcomes of the experiment, called simple events, such that in each trial of the
experiment, one and only one of the results on the list will occur. The set of simple
events for the experiment is called a sample space for the experiment. The sample
space S we have chosen for the die-rolling experiment is

S  {1, 2, 3, 4, 5, 6}

Now consider the outcome, “The number of dots facing up is an even number.”
This outcome is not a simple event, because it will occur whenever 2, 4, or 6 dots
appear, that is, whenever an element in the subset

E  {2, 4, 6}

occurs. Subset E is called a compound event. In general, we have the following definition:

Z DEFINITION 1 Event

Given a sample space S for an experiment, we define an event E to be any

subset of S. If an event E has only one element in it, it is called a simple
event. If event E has more than one element, it is called a compound event.
We say that an event E occurs if any of the simple events in E occurs.
bar67388_ch08_761-779 12/02/06 19:54 Page 762

762 CHAPTER 8 SEQUENCES, INDUCTION, AND PROBABILITY

EXAMPLE 1 Choosing a Sample Space

A nickel and a dime are tossed. How will we identify a sample space for this experiment?

SOLUTIONS

There are a number of possibilities, depending on our interest. We will consider three.

(A) If we are interested in whether each coin falls heads (H) or tails (T), then,
using a tree diagram, we can easily determine an appropriate sample space for
the experiment:

Nickel Dime Combined

Outcomes Outcomes Outcomes
H HH
H
T HT
Start
H TH
T
T TT

Thus,

S1  {HH, HT, TH, TT}

and there are four simple events in the sample space.

(B) If we are interested only in the number of heads that appear on a single toss of
the two coins, then we can let

S2  {0, 1, 2}

and there are three simple events in the sample space.

(C) If we are interested in whether the coins match (M) or don’t match (D), then
we can let

S3  {M, D}

and there are only two simple events in the sample space. 

MATCHED PROBLEM 1

An experiment consists of recording the boy–girl composition of families with two

children.

(A) What is an appropriate sample space if we are interested in the gender of each
child in the order of their births? Draw a tree diagram.
bar67388_ch08_761-779 12/02/06 19:54 Page 763

S E C T I O N 8–5 Sample Spaces and Probability 763

(B) What is an appropriate sample space if we are interested only in the number of
girls in a family?
(C) What is an appropriate sample space if we are interested only in whether the
genders are alike (A) or different (D)?
(D) What is an appropriate sample space for all three interests expressed above? 

In Example 1, sample space S1 contains more information than either S2 or S3.

If we know which outcome has occurred in S1, then we know which outcome has
occurred in S2 and S3. However, the reverse is not true. In this sense, we say that S1
is a more fundamental sample space than either S2 or S3.

Important Remark: There is no one correct sample space for a given

experiment. When specifying a sample space for an experiment, we
include as much detail as necessary to answer all questions of interest
regarding the outcomes of the experiment. If in doubt, include more ele-
ments in the sample space rather than fewer.

Now let’s return to the two-coin problem in Example 1 and the sample space

S1  {HH, HT, TH, TT}

Suppose we are interested in the outcome, “Exactly 1 head is up.” Looking at S1, we
find that it occurs if either of the two simple events HT or TH occurs.* Thus, to say
that the event, “Exactly 1 head is up” occurs is the same as saying the experiment has
an outcome in the set

E  {HT, TH}

This is a subset of the sample space S1. The event E is a compound event.

EXAMPLE 2 Rolling Two Dice

Consider an experiment of rolling two dice. A convenient sample space that will
enable us to answer many questions about interesting events is shown in Figure 1. Let
S be the set of all ordered pairs listed in the figure. Note that the simple event (3, 2)
is to be distinguished from the simple event (2, 3). The former indicates a 3 turned
up on the first die and a 2 on the second, whereas the latter indicates a 2 turned up
on the first die and a 3 on the second. What is the event that corresponds to each of
the following outcomes?

(A) A sum of 7 turns up. (B) A sum of 11 turns up.

(C) A sum less than 4 turns up. (D) A sum of 12 turns up.

*Technically, we should write {HT} and {TH}, because there is a logical distinction between an element
of a set and a subset consisting of only that element. But we will just keep this in mind and drop the
braces for simple events to simplify the notation.
bar67388_ch08_761-779 12/02/06 19:54 Page 764

764 CHAPTER 8 SEQUENCES, INDUCTION, AND PROBABILITY

SECOND DIE

(1, 1) (1, 2) (1, 3) (1, 4) (1, 5) (1, 6)

(2, 1) (2, 2) (2, 3) (2, 4) (2, 5) (2, 6)

FIRST DIE
(3, 1) (3, 2) (3, 3) (3, 4) (3, 5) (3, 6)

(4, 1) (4, 2) (4, 3) (4, 4) (4, 5) (4, 6)

(5, 1) (5, 2) (5, 3) (5, 4) (5, 5) (5, 6)

(6, 1) (6, 2) (6, 3) (6, 4) (6, 5) (6, 6)

Z Figure 1
A sample space for rolling two dice.

SOLUTIONS

(A) By “A sum of 7 turns up,” we mean that the sum of all dots on both turned-up
faces is 7. This outcome corresponds to the event

{(6, 1), (5, 2), (4, 3), (3, 4), (2, 5), (1, 6)}

(B) “A sum of 11 turns up” corresponds to the event

{(6, 5), (5, 6)}

(C) “A sum less than 4 turns up” corresponds to the event

{(1, 1), (2, 1), (1, 2)}

(D) “A sum of 12 turns up” corresponds to the event

{(6, 6)} 

MATCHED PROBLEM 2

Refer to the sample space in Example 2 (Fig. 1). What is the event that corresponds
to each of the following outcomes?

(A) A sum of 5 turns up.

(B) A sum that is a prime number greater than 7 turns up. 

Informally, to facilitate discussion, we often use the terms event and outcome of
an experiment interchangeably. Thus, in Example 2 we might say “the event ‘A sum
of 11 turns up’ ” in place of “the outcome ‘A sum of 11 turns up,’ ” or even write

E  A sum of 11 turns up  {(6, 5), (5, 6)}

bar67388_ch08_761-779 12/02/06 19:54 Page 765

S E C T I O N 8–5 Sample Spaces and Probability 765

Technically speaking, an event is the mathematical counterpart of an outcome of an

experiment.

Z Probability of an Event
The next step in developing our mathematical model for probability studies is the
introduction of a probability function. This is a function that assigns to an arbitrary
event associated with a sample space a real number between 0 and 1, inclusive. We
start by discussing ways in which probabilities are assigned to simple events in S.

Z DEFINITION 2 Probabilities for Simple Events

Given a sample space

S  {e1, e2, . . . , en}

with n simple events, to each simple event ei we assign a real number,

denoted by P(ei), that is called the probability of the event ei. These numbers
may be assigned in an arbitrary manner as long as the following two
conditions are satisfied:
1. 0  P(ei)  1
2. P(e1)  P(e2)  . . .  P(en)  1 The sum of the probabilities of all simple events
in the sample space is 1.

Any probability assignment that meets conditions 1 and 2 is said to be an

acceptable probability assignment.

Our mathematical theory does not explain how acceptable probabilities are
assigned to simple events. These assignments are generally based on the expected or
actual percentage of times a simple event occurs when an experiment is repeated a
large number of times. Assignments based on this principle are called reasonable.
Let an experiment be the flipping of a single coin, and let us choose a sample
space S to be
S  {H, T}

If a coin appears to be fair, we are inclined to assign probabilities to the simple events
in S as follows:

P(H)  12 and P(T)  12

These assignments are based on reasoning that, because there are two ways a coin can
land, in the long run a head will turn up half the time and a tail will turn up half the
time. These probability assignments are acceptable, because both of the conditions for
acceptable probability assignments in Definition 2 are satisfied:

1. 0  P(H)  1, 0  P(T)  1
2. P(H)  P(T)  12  12  1
bar67388_ch08_761-779 12/02/06 19:54 Page 766

766 CHAPTER 8 SEQUENCES, INDUCTION, AND PROBABILITY

But there are other acceptable assignments. Maybe after flipping a coin 1,000
times we find that the head turns up 376 times and the tail turns up 624 times. With
this result, we might suspect that the coin is not fair and assign the simple events in
the sample space S the probabilities

P(H)  .376 and P(T)  .624

This is also an acceptable assignment. But the probability assignment

P(H)  1 and P(T)  0

though acceptable, is not reasonable, unless the coin has two heads. The assignment

P(H)  .6 and P(T)  .8

is not acceptable, because .6  .8  1.4, which violates condition 2 in Definition 2.

In probability studies, the 0 to the left of the decimal is usually omitted. Thus,
we write .8 and not 0.8.
It is important to keep in mind that out of the infinitely many possible accept-
able probability assignments to simple events in a sample space, we are generally
inclined to choose one assignment over another based on reasoning or experimen-
tal results.
Given an acceptable probability assignment for simple events in a sample space S,
how do we define the probability of an arbitrary event E associated with S?

Z DEFINITION 3 Probability of an Event E

Given an acceptable probability assignment for the simple events in a sample

space S, we define the probability of an arbitrary event E, denoted by P(E ),
as follows:
1. If E is the empty set, then P(E)  0.
2. If E is a simple event, then P(E) has already been assigned.
3. If E is a compound event, then P(E) is the sum of the probabilities of all
the simple events in E.
4. If E is the sample space S, then P(E )  P(S)  1. This is a special case
of 3.

EXAMPLE 3 Finding Probabilities of Events

Let’s return to Example 1, the tossing of a nickel and dime, and the sample space

S  {HH, HT, TH, TT}

bar67388_ch08_761-779 12/02/06 19:54 Page 767

S E C T I O N 8–5 Sample Spaces and Probability 767

Because there are four simple outcomes and the coins are assumed to be fair, it
appears that each outcome should occur in the long run 25% of the time. Let’s assign
the same probability of 14 to each simple event in S:

Simple event, ei HH HT TH TT

1 1 1 1
P(ei) 4 4 4 4

This is an acceptable assignment according to Definition 2 and a reasonable assign-

ment for ideal coins that are perfectly balanced or coins close to ideal.

(A) What is the probability of getting exactly one head?

(B) What is the probability of getting at least one head?
(C) What is the probability of getting a head or a tail?
(D) What is the probability of getting three heads?

SOLUTIONS

(A) E1  Getting one head  {HT, TH}

Because E1 is a compound event, we use item 3 in Definition 3 and find P(E1)

by adding the probabilities of the simple events in E1. Thus,
P(E1)  P(HT)  P(TH)  14  14  12

(B) E2  Getting at least 1 head  {HH, HT, TH}

P(E2)  P(HH)  P(HT)  P(TH)
 14  14  14  34

(C) E3  {HH, HT, TH, TT}  S

P(E3)  P(S)  1 1
4  14  14  14  1

(D) E3  Getting three heads   Empty set

P()  0

Z STEPS FOR FINDING PROBABILITIES OF EVENTS

Step 1. Set up an appropriate sample space S for the experiment.

Step 2. Assign acceptable probabilities to the simple events in S.
Step 3. To obtain the probability of an arbitrary event E, add the probabili-
ties of the simple events in E.

The function P defined in steps 2 and 3 is called a probability function. The

domain of this function is all possible events in the sample space S, and the range is
a set of real numbers between 0 and 1, inclusive. 
bar67388_ch08_761-779 12/02/06 19:54 Page 768

768 CHAPTER 8 SEQUENCES, INDUCTION, AND PROBABILITY

MATCHED PROBLEM 3

Return to Matched Problem 1, recording the boy–girl composition of families with

two children and the sample space

S  {BB, BG, GB, GG}

Statistics from the U.S. Census Bureau indicate that an acceptable and reasonable
probability for this sample space is

Simple event, ei BB BG GB GG

P(ei) .26 .25 .25 .24

Find the probabilities of the following events:

(A) E1  Having at least one girl in the family

(B) E2  Having at most one girl in the family
(C) E3  Having two children of the same sex in the family 

Z Equally Likely Assumption

In tossing a nickel and dime (Example 3), we assigned the same probability, 14, to each
simple event in the sample space S  {HH, HT, TH, TT}. By assigning the same
probability to each simple event in S, we are actually making the assumption that each
simple event is as likely to occur as any other. We refer to this as an equally likely
assumption. In general, we have Definition 4.

Z DEFINITION 4 Probability of a Simple Event Under an Equally Likely

Assumption

If, in a sample space

S  {e1, e2, . . . , en}

with n elements, we assume each simple event ei is as likely to occur as any

other, then we assign the probability 1n to each. That is,

1
P(ei) 
n

Under an equally likely assumption, we can develop a very useful formula for
finding probabilities of arbitrary events associated with a sample space S. Consider
the following example.
bar67388_ch08_761-779 12/02/06 19:54 Page 769

S E C T I O N 8–5 Sample Spaces and Probability 769

If a single die is rolled and we assume each face is as likely to come up as any
other, then for the sample space
S  {1, 2, 3, 4, 5, 6}
we assign a probability of 61 to each simple event, because there are six simple events.
Then the probability of
E  Rolling a prime number  {2, 3, 5}
is
P(E)  P(2)  P(3)  P(5)  16  16  16  36  12
Thus, under the assumption that each simple event is as likely to occur as any
other, the computation of the probability of the occurrence of any event E in a sam-
ple space S is the number of elements in E divided by the number of elements in S.

Z THEOREM 1 Probability of an Arbitrary Event Under an Equally

Likely Assumption

If we assume each simple event in sample space S is as likely to occur as

any other, then the probability of an arbitrary event E in S is given by

Number of elements in E n(E )

P(E)  
Number of elements in S n(S )

EXAMPLE 4 Finding Probabilities of Events

If in rolling two dice we assume each simple event in the sample space shown in Fig-
ure 1 on p. 764 is as likely as any other, find the probabilities of the following events:

(A) E1  A sum of 7 turns up (B) E2  A sum of 11 turns up

(C) E3  A sum less than 4 turns up (D) E4  A sum of 12 turns up

SOLUTIONS

Referring to Figure 1, we see that:

n(E1) 6 1 n(E2) 2 1
(A) P(E1)    (B) P(E2)   
n(S) 36 6 n(S) 36 18
n(E3) 3 1 n(E4) 1
(C) P(E3)    (D) P(E4)   
n(S) 36 12 n(S) 36

MATCHED PROBLEM 4

Under the conditions in Example 4, find the probabilities of the following events:

(A) E5  A sum of 5 turns up

(B) E6  A sum that is a prime number greater than 7 turns up 
bar67388_ch08_761-779 12/02/06 19:54 Page 770

770 CHAPTER 8 SEQUENCES, INDUCTION, AND PROBABILITY

ZZZ EXPLORE-DISCUSS 1

A box contains four red balls and seven green balls. A ball is drawn at
random and then, without replacing the first ball, a second ball is drawn. Dis-
cuss whether the equally likely assumption would be appropriate for the
sample space S  {RR, RG, GR, GG}.

We now turn to some examples that make use of the counting techniques devel-
oped in Section 8-4.

EXAMPLE 5 Drawing Cards

In drawing 5 cards from a 52-card deck without replacement, what is the probability
of getting five spades?

SOLUTION

Let the sample space S be the set of all 5-card hands from a 52-card deck. Because
the order in a hand does not matter, n(S )  C52,5. The event we seek is

E  Set of all 5-card hands from 13 spades

Again, the order does not matter and n(E )  C13,5. Thus, assuming each 5-card hand
is as likely as any other,

n(E ) C13,5 13!5!8! 13! 5!47!

P(E )       .0005 
n(S ) C52,5 52!5!47! 5!8! 52!

MATCHED PROBLEM 5

In drawing 7 cards from a 52-card deck without replacement, what is the probability
of getting seven hearts? 

EXAMPLE 6 Selecting Committees

The board of regents of a university is made up of 12 men and 16 women. If a com-

mittee of six is chosen at random, what is the probability that it will contain three
men and three women?

SOLUTION

Let S  Set of all 6-person committees out of 28 people:

n(S )  C28,6
bar67388_ch08_761-779 12/02/06 19:54 Page 771

S E C T I O N 8–5 Sample Spaces and Probability 771

Let E  Set of all 6-person committees with 3 men and 3 women. To find n(E ), we
use the multiplication principle and the following two operations:

O1: Select 3 men out of the 12 available N1: C12,3

O2: Select 3 women out of the 16 available N2: C16,3

Thus,
n(E)  C12,3  C16,3
and
n(E) C12,3  C16,3
P(E)    .327 
n(S) C28,6

MATCHED PROBLEM 6

What is the probability that the committee in Example 6 will have four men and two
women? 

Z Empirical Probability
In the earlier examples in Section 8-5, we made a reasonable assumption about an exper-
iment and used deductive reasoning to assign probabilities. For example, it is reasonable
to assume that an ordinary coin will come up heads about as often as it will come up
tails. Probabilities determined in this manner are called theoretical probabilities. No
experiments are ever conducted. But what if the theoretical probabilities are not obvious?
Then we assign probabilities to simple events based on the results of actual experiments.
Probabilities determined from the results of actually performing an experiment are called
empirical probabilities. As an experiment is repeated over and over, the percentage of
times an event occurs may get closer and closer to a single fixed number. If so, this sin-
gle fixed number is generally called the actual probability of the event.

ZZZ EXPLORE-DISCUSS 2

Like a coin, a thumbtack tossed into the air will land in one of two positions,
point up or point down [Fig. 2(a)]. Unlike a coin, we would not expect both
events to occur with the same frequency. Indeed, the frequencies of landing point
up and point down may well vary from one thumbtack to another [Fig. 2(b)].
Find two thumbtacks of different sizes and guess which one is likely to land
point up more frequently. Then toss each tack 100 times and record the num-
ber of times each lands point up. Did the experiment confirm your initial guess?

(a) Point up or point down (b) Two different tacks

Z Figure 2
bar67388_ch08_761-779 12/02/06 19:54 Page 772

772 CHAPTER 8 SEQUENCES, INDUCTION, AND PROBABILITY

Suppose when tossing one of the thumbtacks in Explore-Discuss 2, we observe

that the tack lands point up 43 times and point down 57 times. Based on this exper-
iment, it seems reasonable to say that for this particular thumbtack

43
P(Point up)   .43
100
57
P(Point down)   .57
100

Probability assignments based on the results of repeated trials of an experiment are

called approximate empirical probabilities.
In general, if we conduct an experiment n times and an event E occurs with fre-
quency f(E ), then the ratio f(E )n is called the relative frequency of the occurrence
of event E in n trials. We define the empirical probability of E, denoted by P(E ),
by the number, if it exists, that the relative frequency f(E )n approaches as n gets
larger and larger. Of course, for any particular n, the relative frequency f(E )n is
generally only approximately equal to P(E ). However, as n increases, we expect the
approximation to improve.

Z DEFINITION 5 Empirical Probability

If f(E) is the frequency of event E in n trials, then

Frequency of occurrence of E f (E )
P(E )  
Total number of trials n

If we can also deduce theoretical probabilities for an experiment, then we expect

the approximate empirical probabilities to approach the theoretical probabilities. If this
does not happen, then we should begin to suspect the manner in which the theoreti-
cal probabilities were computed. If P(E ) is the theoretical probability of an event E and
the experiment is performed n times, then the expected frequency of the occurrence
of E is n  P(E ).

EXAMPLE 7 Finding Approximate Empirical and Theoretical Probabilities

Two coins are tossed 500 times with the following frequencies of outcomes:

Two heads: 121

One head: 262
Zero heads: 117

(A) Compute the approximate empirical probability for each outcome.

(B) Compute the theoretical probability for each outcome.
(C) Compute the expected frequency for each outcome.
bar67388_ch08_761-779 12/02/06 19:54 Page 773

S E C T I O N 8–5 Sample Spaces and Probability 773

SOLUTIONS

121
(A) P(two heads)   .242
500
262
P(one head)   .524
500
117
P(zero heads)   .234
500

(B) A sample space of equally likely simple events is S  {HH, HT, TH, TT}. Let

E1  two heads  5HH6

E2  one head  5HT, TH6
E3  zero heads  5TT6

Then

n(E1) 1
P(E1)    .25
n(S) 4
n(E2) 2
P(E2)    .50
n(S) 4
n(E3) 1
P(E3)    .25
n(S) 4

(C) The expected frequencies are

E1: 500(.25)  125

E2: 500(.5)  250
E3: 500(.25)  125

The actual frequencies obtained from performing the experiment are reasonably
close to the expected frequencies. Increasing the number of trials of the experi-
ment would produce even better approximations. 

MATCHED PROBLEM 7

One die is rolled 500 times with the following frequencies of outcomes:

Outcome 1 2 3 4 5 6

Frequency 89 83 77 91 72 88

(A) Compute the approximate empirical probability for each outcome.

(B) Compute the theoretical probability for each outcome.
(C) Compute the expected frequency for each outcome. 
bar67388_ch08_761-779 12/4/06 11:25 Page 774

774 CHAPTER 8 SEQUENCES, INDUCTION, AND PROBABILITY

Adding two such sequences together produces a

sequence of 500 terms in which each term repre-
sents the number of heads in a toss of two coins
Technology [see Fig. 3(a)]. We determine the frequency of each
Connections outcome (0, 1, or 2 heads) in 500 tosses of two
coins as follows: first, we construct a histogram
[Figs. 3(b) and 3(c)], then we use the TRACE com-
The data in Example 7 were not generated by mand to read off the frequencies [Figs. 3(d), 3(e),
tossing two coins 500 times. Instead, the experi- and 3(f)]. Compare with the data of Example 7.
ment was simulated by a random number gen- If you perform the same simulation on your
erator on a graphing calculator. The command graphing calculator, you are not likely to get exactly
randInt (0, 1, 500) produces a random sequence the same results. But the approximate empirical
of 500 terms; each term is 0 or 1 with equal likli- probabilities you obtain will be close to the theo-
hood. Thinking of 1 as heads and 0 as tails, such a retical probabilities.
sequence represents 500 tosses of a single coin.

(a) Generating the (b) Setting up the (c) Selecting the

random numbers histogram window variables

(d) 0 heads: 117 (e) 1 head: 262 (f) 2 heads: 121

Z Figure 3 Simulating 500 tosses of two coins.

EXAMPLE 8 Empirical Probabilities for an Insurance Company

An insurance company selected 1,000 drivers at random in a particular city to deter-

mine a relationship between age and accidents. The data obtained are listed in Table 1.
Compute the approximate empirical probabilities of the following events for a driver
chosen at random in the city:
bar67388_ch08_761-779 12/02/06 19:54 Page 775

S E C T I O N 8–5 Sample Spaces and Probability 775

(A) E1: being under 20 years old and having exactly three accidents in 1 year
(B) E2: being 30–39 years old and having one or more accidents in 1 year
(C) E3: having no accidents in 1 year
(D) E4: being under 20 years old or* having exactly three accidents in 1 year

Table 1
Accidents in 1 Year

Age 0 1 2 3 Over 3

Under 20 50 62 53 35 20
20–29 64 93 67 40 36
30–39 82 68 32 14 4
40–49 38 32 20 7 3
Over 49 43 50 35 28 24

SOLUTIONS

35
(A) P(E1)   .035
1,000
68  32  14  4
(B) P(E2)   .118
1,000
50  64  82  38  43
(C) P(E3)   .277
1,000
50  62  53  35  20  40  14  7  28
(D) P(E4)   .309
1,000

Notice that in this type of problem, which is typical of many realistic problems,
approximate empirical probabilities are the only type we can compute. 

MATCHED PROBLEM 8

Referring to Table 1 in Example 8, compute the approximate empirical probabilities

of the following events for a driver chosen at random in the city:

(A) E1: being under 20 years old with no accidents in 1 year

(B) E2: being 20–29 years old and having fewer than two accidents in 1 year
(C) E3: not being over 49 years old 

*Interpret “or” in its inclusive sense, as customary in mathematics (a driver who is both under 20 and
has three accidents must be counted once in the frequency of E4).
bar67388_ch08_761-779 12/02/06 19:54 Page 776

776 CHAPTER 8 SEQUENCES, INDUCTION, AND PROBABILITY

Approximate empirical probabilities are often used to test theoretical probabili-

ties. Equally likely assumptions may not be justified in reality. In addition to this use,
there are many situations in which it is either very difficult or impossible to compute
the theoretical probabilities for given events. For example, insurance companies use
past experience to establish approximate empirical probabilities to predict future acci-
dent rates; baseball teams use batting averages, which are approximate empirical prob-
abilities based on past experience, to predict the future performance of a player; and
pollsters use approximate empirical probabilities to predict outcomes of elections.

ANSWERS TO MATCHED PROBLEMS

1. (A) S1  {BB, BG, GB, GG}; Sex of Sex of Combined

First Child Second Child Outcomes
B BB
B
G BG
B GB
G
G GG
(B) S2  {0, 1, 2} (C) S3  {A, D} (D) The sample space in part A.
2. (A) {(4, 1), (3, 2), (2, 3), (1, 4)} (B) {(6, 5), (5, 6)}
3. (A) .74 (B) .76 (C) .5 4. (A) P(E5)  19 (B) P(E6)  181
5. C13,7C52,7  .000013 6. C12,4  C16,2C28,6  .158
7. (A) P(E1)  .178, P(E2)  .166, P(E3)  .154, P(E4)  .182, P(E5)  .144, P(E6)  .176
(B) 16  .167 for each (C) 83.3 for each
8. (A) P(E1)  .05 (B) P(E2)  .157 (C) P(E3)  .82

8-5 Exercises

1. A single fair die is rolled. What is the probability of getting 8. Two fair dice are rolled. What is the probability of getting
a one or a six? double sixes?
2. A single fair die is rolled. What is the probability of getting
a number greater than three?
3. A single card is drawn from a standard 52-card deck. What
is the probability of getting a red card?
4. A single card is drawn from a standard 52-card deck. What
is the probability of getting a club?
5. A fair coin is tossed twice. What is the probability of
getting two heads?
6. A fair coin is tossed twice. What is the probability of get-
ting at least one head?
7. Two fair dice are rolled. What is the probability of getting
doubles?
bar67388_ch08_761-779 12/02/06 19:54 Page 777

S E C T I O N 8–5 Sample Spaces and Probability 777

9. A single card is drawn from a standard 52-card deck. What 24. A combination lock has five wheels, each labeled with the
is the probability of getting a king or a queen? 10 digits from 0 to 9. If an opening combination is a partic-
ular sequence of five digits with no repeats, what is the
10. A single card is drawn from a standard 52-card deck. What
probability of a person guessing the right combination?
is the probability of getting a numbered card (that is, a two
through ten)?
Problems 25–30 involve an experiment consisting of dealing
11. A fair coin is tossed three times? What is the probability of
5 cards from a standard 52-card deck. In Problems 25–28,
getting exactly two tails?
what is the probability of being dealt:
12. A fair coin is tossed three times? What is the probability of
25. Five black cards
getting three tails?
26. Five hearts
13. How would you interpret P(E)  1?
27. Five face cards if an ace is considered to be a face card.
14. How would you interpret P(E)  0?
28. Five nonface cards if an ace is considered to be a one and
15. A spinner can land on four different colors: red (R), green
not a face.
(G), yellow (Y), and blue (B). If we do not assume each color
is as likely to turn up as any other, which of the following 29. If we are interested in the number of aces in a 5-card hand,
probability assignments have to be rejected, and why? would S  {0, 1, 2, 3, 4} be an acceptable sample space?
(A) P(R)  .15, P(G)  .35, P(Y )  .50, P(B)  .70 Would it be an equally-likely sample space? Explain.
(B) P(R)  .32, P(G)  .28, P(Y )  .24, P(B)  .30
(C) P(R)  .26, P(G)  .14, P(Y )  .30, P(B)  .30 30. If we are interested in the number of black cards in a 5-card
hand, would S  {0, 1, 2, 3, 4, 5} be an acceptable sample
16. Under the probability assignments in Problem 15, part C, space? Would it be an equally-likely sample space? Explain.
what is the probability that the spinner will not land on blue?
17. Under the probability assignments in Problem 15, part C, 31. If four-digit numbers less than 5,000 are randomly formed
what is the probability that the spinner will land on red or from the digits 1, 3, 5, 7, and 9, what is the probability of
yellow? forming a number divisible by 5? Digits may be repeated;
for example, 1,355 is acceptable.
18. Under the probability assignments in Problem 15, part C,
what is the probability that the spinner will not land on red 32. If code words of four letters are generated at random using the
or yellow? letters A, B, C, D, E, and F, what is the probability of forming
a word without a vowel in it? Letters may be repeated.
19. A ski jumper has jumped over 300 feet in 25 out of 250
jumps. What is the approximate empirical probability of the 33. Suppose five thank-you notes are written and five envelopes
next jump being over 300 feet? are addressed. Accidentally, the notes are randomly in-
serted into the envelopes and mailed without checking the
20. In a certain city there are 4,000 youths between 16 and 20
addresses. What is the probability that all five notes will be
years old who drive cars. If 560 of them were involved in
inserted into the correct envelopes?
accidents last year, what is the approximate empirical prob-
ability of a youth in this age group being involved in an ac- 34. Suppose six people check their coats in a checkroom. If all
cident this year? claim checks are lost and the six coats are randomly re-
turned, what is the probability that all six people will get
21. Out of 420 times at bat, a baseball player gets 189 hits.
their own coats back?
What is the approximate empirical probability that the
player will get a hit next time at bat?
An experiment consists of rolling two fair dice and adding the
22. In a medical experiment, a new drug is found to help 2,400
dots on the two sides facing up. Using the sample space shown
out of 3,000 people. If a doctor prescribes the drug for a
in Figure 1 (p. 764) and assuming each simple event is as likely
particular patient, what is the approximate empirical proba-
as any other, find the probabilities of the sums of dots indicated
bility that the patient will be helped?
in Problems 35–50.
23. A small combination lock on a suitcase has three wheels,
35. Sum is 2. 36. Sum is 10.
each labeled with the 10 digits from 0 to 9. If an opening
combination is a particular sequence of three digits with no 37. Sum is 6. 38. Sum is 8.
repeats, what is the probability of a person guessing the
right combination? 39. Sum is less than 5. 40. Sum is greater than 8.
41. Sum is not 7 or 11. 42. Sum is not 2, 4, or 6.
bar67388_ch08_761-779 12/02/06 19:54 Page 778

778 CHAPTER 8 SEQUENCES, INDUCTION, AND PROBABILITY

43. Sum is 1. 44. Sum is not 13. 55. (A) Is it possible to get 29 heads in 30 flips of a fair coin?
Explain.
45. Sum is divisible by 3. 46. Sum is divisible by 4.
(B) If you flip a coin 50 times and get 42 heads, would you
47. Sum is 7 or 11 (a “natural”). suspect that the coin was unfair? Why or why not? If you
suspect an unfair coin, what empirical probabilities would
48. Sum is 2, 3, or 12 (“craps”). you assign to the simple events of the sample space?
49. Sum is divisible by 2 or 3. 56. (A) Is it possible to get nine double sixes in 12 rolls of a pair
50. Sum is divisible by 2 and 3. of fair dice? Explain.
(B) If you roll a pair of dice 40 times and get 14 double
51. Five thousand people work in a large auto plant. An indi- sixes, would you suspect that the dice were unfair? Why
vidual is selected at random and his or her birthday (month or why not? If you suspect loaded dice, what empirical
and day, not year) is recorded. Set up an appropriate sample probability would you assign to the event of rolling a
space for this experiment and assign acceptable probabili- double six?
ties to the simple events.
52. In a hotly contested three-way race for governor of Min- An experiment consists of tossing three fair coins, but one of the
nesota, the leading candidates are running neck-and-neck three coins has a head on both sides. Compute the probabilities
while the third candidate is receiving half the support of either of obtaining the indicated results in Problems 57–62.
of the others. Registered voters are chosen at random and are
asked for which of the three they are most likely to vote. Set 57. One head 58. Two heads
up an appropriate sample space for the random survey exper-
59. Three heads 60. Zero heads
iment and assign acceptable probabilities to the simple events.
61. More than one head 62. More than one tail
53. A pair of dice is rolled 500 times with the following fre-
quencies:
An experiment consists of rolling two fair dice and adding the
Sum 2 3 4 5 6 7 8 9 10 11 12
dots on the two sides facing up. Each die has one dot on two
Frequency 11 35 44 50 71 89 72 52 36 26 14 opposite faces, two dots on two opposite faces, and three dots
(A) Compute the approximate empirical probability for on two opposite faces. Compute the probabilities of obtaining
each outcome. the indicated sums in Problems 63–70.
(B) Compute the theoretical probability for each outcome, 63. 2 64. 3
assuming fair dice.
(C) Compute the expected frequency of each outcome. 65. 4 66. 5
(D) Describe how a random number generator could be
67. 6 68. 7
used to simulate this experiment. If your graphing
calculator has a random number generator, use it to 69. An odd sum 70. An even sum
simulate 500 tosses of a pair of dice and compare your
results with part C.
An experiment consists of dealing 5 cards from a standard
54. Three coins are flipped 500 times with the following fre- 52-card deck. In Problems 71–78, what is the probability of
quencies of outcomes: being dealt the following cards?
Three heads: 58 Two heads: 198 71. Five cards, jacks through aces
One head: 190 Zero heads: 54
72. Five cards, 2 through 10
(A) Compute the approximate empirical probability for
each outcome. 73. Four aces
(B) Compute the theoretical probability for each outcome, 74. Four of a kind
assuming fair coins.
(C) Compute the expected frequency of each outcome. 75. Straight flush, ace high; that is, 10, jack, queen, king, ace in
(D) Describe how a random number generator could be one suit
used to simulate this experiment. If your graphing cal- 76. Straight flush, starting with 2; that is, 2, 3, 4, 5, 6 in one suit
culator has a random number generator, use it to simu-
late 500 tosses of three coins and compare your results 77. Two aces and three queens
with part C. 78. Two kings and three aces
bar67388_ch08_761-779 12/02/06 19:54 Page 779

S E C T I O N 8–6 Binomial Formula 779

APPLICATIONS Compute the approximate empirical probabilities:

(A) Of a household earning $12,000–$19,999 per year and
79. MARKET ANALYSIS A company selected 1,000 households owning exactly three television sets
at random and surveyed them to determine a relationship be- (B) Of a household earning $20,000–$39,999 per year and
tween income level and the number of television sets in a home. owning more than one television set
The information gathered is listed in the table: (C) Of a household earning $60,000 or more per year or owning
Televisions per Household more than three television sets
(D) Of a household not owning zero television sets
Yearly Income ($) 0 1 2 3 Above 3
80. MARKET ANALYSIS Use the sample results in Problem 79
to compute the approximate empirical probabilities:
Less than 12,000 0 40 51 11 0
(A) Of a household earning $40,000–$59,999 per year and
12,000–19,999 0 70 80 15 1 owning zero television sets
(B) Of a household earning $12,000–$39,999 per year and
20,000–39,999 2 112 130 80 12 owning more than two television sets
40,000–59,999 10 90 80 60 21 (C) Of a household earning less than $20,000 per year or
owning exactly two television sets
60,000 or more 30 32 28 25 20 (D) Of a household not owning more than three television sets

8-6 Binomial Formula

Z Pascal’s Triangle
Z The Binomial Formula
Z Proof of the Binomial Formula

The binomial form

(a  b)n

where n is a natural number, appears more frequently than you might expect. It turns
out that the coefficients in the expansion are related to probability concepts that we
have already discussed.

Z Pascal’s Triangle
Let’s begin by expanding (a  b)n for the first few values of n. We include n  0,
which is not a natural number, for reasons of completeness that will become apparent
later.

(a  b)0  1
(a  b)1  a  b (1)
(a  b)2  a2  2ab  b2
(a  b)3  a3  3a2b  3ab2  b3