426 CHAPTER 8 Probability

69. Explain how the three events A, B, and C from a sample 78. Consider the command in Figure A and the associated statis-
space S are related to each other in order for the following tical plot in Figure B.
equation to hold true:
(A) Explain why the command does not simulate 50 repeti-
P1A ∪ B ∪ C2 = P1A2 + P1B2 + P1C2 - P1A ¨ B2 tions of rolling a pair of dice and recording their sum.
70. Explain how the three events A, B, and C from a sample (B) Describe an experiment that is simulated by this command.
space S are related to each other in order for the following
equation to hold true: (C) Simulate 200 repetitions of the experiment you
described in part (B), and find the empirical probability
P1A ∪ B ∪ C2 = P1A2 + P1B2 + P1C2
of recording a 7 or 8.
71. Show that the solution to the birthday problem in Example 5
can be written in the form (D) What is the theoretical probability of recording a 7 or 8?

365Pn
12
P1E2 = 1 - n
365
For a calculator that has a nPr function, explain why this form
may be better for direct evaluation than the other form used 4 14
in the solution to Example 5. Try direct evaluation of both
forms on a calculator for n = 25.
72. Many (but not all) calculators experience an overflow error 0
when computing 365Pn for n 7 39 and when computing 365n. (A) (B)
Explain how you would evaluate P1E2 for any n 7 39 on
such a calculator.
73. In a group of n people 1n … 122, what is the probability that
at least 2 of them have the same birth month? (Assume that Applications
any birth month is as likely as any other.)
74. In a group of n people 1n … 1002, each person is asked to 79. Market research. From a survey involving 1,000 university
select a number between 1 and 100, write the number on a students, a market research company found that 750 students
slip of paper and place the slip in a hat. What is the probabil- owned laptops, 450 owned cars, and 350 owned cars and lap-
ity that at least 2 of the slips in the hat have the same number tops. If a university student is selected at random, what is the
written on them? (empirical) probability that
75. If the odds in favor of an event E occurring are a to b, show that (A) The student owns either a car or a laptop?
a (B) The student owns neither a car nor a laptop?
P1E2 =
a + b 80. Market research. Refer to Problem 79. If a university student
[Hint: Solve the equation P1E2 >P1E′2 = a>b for P1E2.] is selected at random, what is the (empirical) probability that
76. If P1E2 = c>d, show that odds in favor of E occurring are (A) The student does not own a car?
c to d - c.
(B) The student owns a car but not a laptop?
77. The command in Figure A was used on a graphing calculator
to simulate 50 repetitions of rolling a pair of dice and recording 81. Insurance. By examining the past driving records of city
their sum. A statistical plot of the results is shown in Figure B. drivers, an insurance company has determined the following
(empirical) probabilities:
(A) Use Figure B to find the empirical probability of rolling
a 7 or 8. Miles Driven per Year
(B) What is the theoretical probability of rolling a 7 or 8? Less 10,000– More
than 15,000, than
(C) Use a graphing calculator to simulate 200 repetitions of Totals
10,000, inclusive, 15,000,
rolling a pair of dice and recording their sum, and find M1 M2 M3
the empirical probability of rolling a 7 or 8.
Accident A .05 .1 .15 .3
14 No Accident A′ .15 .2 .35 .7
Totals .2 .3 .5 1.0

4 14 If a city driver is selected at random, what is the probability

that
(A) He or she drives less than 10,000 miles per year or has
0 an accident?
(A) (B) (B) He or she drives 10,000 or more miles per year and has
no accidents?
 SECTION 8.3 Conditional Probability, Intersection, and Independence 427

82. Insurance. Use the (empirical) probabilities in Problem 81 their preferences in an upcoming election. (In the table,
to find the probability that a city driver selected at random D = Democrat, R = Republican, and U = Unaffiliated.)
(A) Drives more than 15,000 miles per year or has an accident D R U Totals
(B) Drives 15,000 or fewer miles per year and has an accident Candidate A A 200 100 85 385
Candidate B B 250 230 50 530
83. Quality control. A shipment of 60 game players, including
9 that are defective, is sent to a retail store. The receiving No Preference N 50 20 15 85
department selects 10 at random for testing and rejects the Totals 500 350 150 1,000
whole shipment if 1 or more in the sample are found to be
defective. What is the probability that the shipment will be 87. Politics. If a state resident is selected at random, what is the
rejected? (empirical) probability that the resident is

84. Quality control. An assembly plant produces 40 outboard (A) Not affiliated with a political party or has no preference?
motors, including 7 that are defective. The quality control What are the odds for this event?
department selects 10 at random (from the 40 produced) for (B) Affiliated with a political party and prefers candidate A?
testing and will shut down the plant for trouble shooting if 1 What are the odds against this event?
or more in the sample are found to be defective. What is the
probability that the plant will be shut down? 88. Politics. If a state resident is selected at random, what is the
(empirical) probability that the resident is
85. Medicine. In order to test a new drug for adverse reactions,
the drug was administered to 1,000 test subjects with the fol- (A) A Democrat or prefers candidate B? What are the odds
lowing results: 60 subjects reported that their only adverse for this event?
reaction was a loss of appetite, 90 subjects reported that their (B) Not a Democrat and has no preference? What are the
only adverse reaction was a loss of sleep, and 800 subjects odds against this event?
reported no adverse reactions at all. If this drug is released
for general use, what is the (empirical) probability that a Answers to Matched Problems
person using the drug will suffer both a loss of appetite and a
1 2 1 7
loss of sleep? 1. (A) (B) 2. (A) (B)
3 3 12 18
86. Product testing. To test a new car, an automobile manufac-
47
turer wants to select 4 employees to test drive the car for 1 3. ≈ .336 4. .92 5. .016
year. If 12 management and 8 union employees volunteer to 140
be test drivers and the selection is made at random, what is the 5
6. (A) 5 : 31 (B) $31 7. ≈ .455
probability that at least 1 union employee is selected? 11
1
Problems 87 and 88 refer to the data in the following table, 8. (A) P1D′ ¨ R′2 = .1; odds for D′ ¨ R′ = or 1 : 9
9
obtained from a random survey of 1,000 residents of a state. 2
The participants were asked their political affiliations and (B) P1D ∪ R′2 = .6; odds against D ∪ R′ = or 2 : 3
3

8.3 Conditional Probability, Intersection, and Independence

• Conditional Probability In Section 8.2, we learned that the probability of the union of two events is related to
• Intersection of Events:
the sum of the probabilities of the individual events (Theorem 1, p. 416):
Product Rule P1A ∪ B2 = P1A2 + P1B2 - P1A ¨ B2
• Probability Trees In this section, we will learn how the probability of the intersection of two events is
• Independent Events related to the product of the probabilities of the individual events. But first we must
investigate the concept of conditional probability.
• Summary

Conditional Probability
The probability of an event may change if we are told of the occurrence of another
event. For example, if an adult (21 years or older) is selected at random from all
adults in the United States, the probability of that person having lung cancer would
not be high. However, if we are told that the person is also a heavy smoker, we would
want to revise the probability upward.
In general, the probability of the occurrence of an event A, given the occurrence
of another event B, is called a conditional probability and is denoted by P1A∙ B2.
428 CHAPTER 8 Probability

In the preceding situation, events A and B would be

A = adult has lung cancer
B = adult is a heavy smoker
and P1A∙ B2 would represent the probability of an adult having lung cancer, given
that he or she is a heavy smoker.
Our objective now is to try to formulate a precise definition of P1A∙ B2. It is
helpful to start with a relatively simple problem, solve it intuitively, and then general-
ize from this experience.
What is the probability of rolling a prime number (2, 3, or 5) in a single roll of a
1 A 4
fair die? Let
3 6 S = 51, 2, 3, 4, 5, 66
2 5 Then the event of rolling a prime number is (Fig. 1)
A = 52, 3, 56
Figure 1 Thus, since we assume that each simple event in the sample space is equally likely,
n1A2 3 1
P1A2 = = =
n1S2 6 2
Now suppose you are asked, “In a single roll of a fair die, what is the probability that
B a prime number has turned up if we are given the additional information that an odd num-
1 A 4
ber has turned up?” The additional knowledge that another event has occurred, namely,
3 6 B = odd number turns up
2 5 puts the problem in a new light. We are now interested only in the part of event A
(rolling a prime number) that is in event B (rolling an odd number). Event B, since we
know it has occurred, becomes the new sample space. The Venn diagrams in Figure 2
illustrate the various relationships. Thus, the probability of A given B is the number of
1 A elements in B divided by the total number of elements in B. Symbolically,
3 n1A ¨ B2
AB 2
5 P1A∙ B2 = =
n1B2 3
Dividing the numerator and denominator of n1A ¨ B2 >n1B2 by n(S), the number of
Figure 2 B is the new sample space.
elements in the original sample space, we can express P1A∙ B2 in terms of P1A ¨ B2
and P(B):*
n1A ¨ B2
n1A ¨ B2 n1S2 P1A ¨ B2
P1A∙ B2 = = =
n1B2 n1B2 P1B2
n1S2
Using the right side to compute P1A∙ B2 for the preceding example, we obtain the
same result: 2
P1A ¨ B2 6 2
P1A∙ B2 = = 3
=
P1B2 6
3
We use the formula above to motivate the following definition of conditional
probability, which applies to any sample space, including those having simple events
that are not equally likely (see Example 1).

Definition Conditional Probability

For events A and B in an arbitrary sample space S, we define the conditional prob-
ability of A given B by
P1A ¨ B2
P1A∙ B2 = P1B2 ≠ 0 (1)
P1B2

*Note that P1A∙ B2 is a probability based on the new sample space B, while P1A ¨ B2 and P(B) are both
probabilities based on the original sample space S.
 SECTION 8.3 Conditional Probability, Intersection, and Independence 429

ExamplE 1 Conditional Probability A pointer is spun once on a circular

spinner (Fig. 3). The probability assigned to the pointer landing on a given integer
(from 1 to 6) is the ratio of the area of the corresponding circular sector to the area
of the whole circle, as given in the table:
ei 1 2 3 4 5 6
S = 51, 2, 3, 4, 5, 66
P1ei 2 .1 .2 .1 .1 .3 .2

(A) What is the probability of the pointer landing on a prime number?

(B) What is the probability of the pointer landing on a prime number, given that it
Figure 3
landed on an odd number?
Solution Let the events E and F be defined as follows:

E = pointer lands on a prime number = 52, 3, 56

F = pointer lands on an odd number = 51, 3, 56
(A) P1E2 = P122 + P132 + P152
= .2 + .1 + .3 = .6
(B) First note that E ¨ F = 53, 56.
P1E ¨ F2 P132 + P152
P1E ∙ F2 = =
P1F2 P112 + P132 + P152
.1 + .3 .4
= = = .8
.1 + .1 + .3 .5

Matched Problem 1 Refer to Example 1.

(A) What is the probability of the pointer landing on a number greater than 4?
(B) What is the probability of the pointer landing on a number greater than 4,
given that it landed on an even number?

ExamplE 2 Safety Research Suppose that city records produced the follow-
ing probability data on a driver being in an accident on the last day of a Memorial
Day weekend:
Accident No Accident
A A= Totals
Rain R .025 .335 .360
No Rain R′ .015 .625 .640 S = 5RA, RA′, R′A, R′A′6
Totals .040 .960 1.000

(A) Find the probability of an accident, rain or no rain.

(B) Find the probability of rain, accident or no accident.
(C) Find the probability of an accident and rain.
(D) Find the probability of an accident, given rain.
Solution
(A) Let A = 5RA, R′A6 Event: “accident”

P1A2 = P1RA2 + P1R′A2 = .025 + .015 = .040

(B) Let R = 5RA, RA′6 Event: “rain”

P1R2 = P1RA2 + P1RA′2 = .025 + .335 = .360

430 CHAPTER 8 Probability

(C) A ¨ R = 5RA6 Event: “accident and rain”

P1A ¨ R2 = P1RA2 = .025

P1A ¨ R2 .025
(D) P1A∙ R2 = = = .069 Event: “accident, given rain”
P1R2 .360
Compare the result in part (D) with that in part (A). Note that P1A∙ R2 ≠ P1A2,
and the probability of an accident, given rain, is higher than the probability of an
accident without the knowledge of rain.

Matched Problem 2 Referring to the table in Example 2, determine the following:

(A) Probability of no rain
(B) Probability of an accident and no rain
(C) Probability of an accident, given no rain [Use formula (1) and the results of
parts (A) and (B).]

Intersection of Events: Product Rule

Let’s return to the original problem of this section, that is, representing the prob-
ability of an intersection of two events in terms of the probabilities of the individual
events. If P1A2 ≠ 0 and P1B2 ≠ 0, then using formula (1), we can write
P1A ¨ B2 P1B ¨ A2
P1A∙ B2 = and P1B∙ A2 =
P1B2 P1A2
Solving the first equation for P1A ¨ B2 and the second equation for P1B ¨ A2,
we have
P1A ¨ B2 = P1B2P1A∙ B2 and P1B ¨ A2 = P1A2P1B∙ A2
Since A ¨ B = B ¨ A for any sets A and B, it follows that
P1A ¨ B2 = P1B2P1A∙ B2 = P1A2P1B∙ A2
and we have the product rule:

theorem 1 Product Rule

For events A and B with nonzero probabilities in a sample space S,
P1A ¨ B2 = P1A2P1B∙ A2 = P1B2P1A∙ B2 (2)

and we can use either P1A2P1B∙ A2 or P1B2P1A∙ B2 to compute P1A ¨ B2.

ExamplE 3 Consumer Survey If 60% of a department store’s customers are

female and 75% of the female customers have credit cards at the store, what is the
probability that a customer selected at random is a female and has a store credit card?
Solution Let
S = all store customers
F = female customers
C = customers with a store credit card
If 60% of the customers are female, then the probability that a customer selected at
random is a female is
P1F2 = .60
Since 75% of the female customers have store credit cards, the probability that a cus-
tomer has a store credit card, given that the customer is a female, is
P1C∙ F2 = .75
 SECTION 8.3 Conditional Probability, Intersection, and Independence 431

Using equation (2), the probability that a customer is a female and has a store credit
card is
P1F ¨ C2 = P1F2P1C∙ F2 = 1.6021.752 = .45

Matched Problem 3 If 80% of the male customers of the department store in

Example 3 have store credit cards, what is the probability that a customer selected
at random is a male and has a store credit card?

Probability Trees
We used tree diagrams in Section 7.3 to help us count the number of combined out-
comes in a sequence of experiments. In a similar way, we will use probability trees to
help us compute the probabilities of combined outcomes in a sequence of experiments.

ExamplE 4 Probability Tree Two balls are drawn in succession, without

replacement, from a box containing 3 blue and 2 white balls (Fig. 4). What is the
probability of drawing a white ball on the second draw?
Solution We start with a tree diagram (Fig. 5) showing the combined outcomes
of the two experiments (first draw and second draw). Then we assign a probability
to each branch of the tree (Fig. 6). For example, we assign the probability 25 to the
branch Sw1, since this is the probability of drawing a white ball on the first draw
Figure 4
(there are 2 white balls and 3 blue balls in the box). What probability should be
assigned to the branch w1w2? This is the conditional probability P1w2 ∙ w1 2, that
is, the probability of drawing a white ball on the second draw given that a white
FIRST SECOND ball was drawn on the first draw and not replaced. Since the box now contains 1
DRAW DRAW white ball and 3 blue balls, the probability is 14. Continuing in the same way, we
w™ assign probabilities to the other branches of the tree and obtain Figure 6.
w¡ What is the probability of the combined outcome w1 ¨ w2, that is, the
b™
S
Start b¡
w™ probability of drawing a white ball on the first draw and a white ball on the second
b™ draw?* Using the product rule (2), we have
Figure 5 P1w1 ¨ w2 2 = P1w1 2P1w2 ∙ w1 2
2 1 1
FIRST SECOND = a ba b =
5 4 10
DRAW DRAW
1
4 w™ The combined outcome w1 ¨ w2 corresponds to the unique path Sw1w2 in the tree
2 w¡
diagram, and we see that the probability of reaching w2 along this path is the product
5
3 of the probabilities assigned to the branches on the path. Reasoning in this way, we
b™
S 4
2 obtain the probability of each remaining combined outcome by multiplying the prob-
w™
3
4
abilities assigned to the branches on the path corresponding to the given combined
5 b¡ outcomes. These probabilities are often written at the ends of the paths to which they
2
4
b™ correspond (Fig. 7).
Figure 6 FIRST SECOND
DRAW DRAW
2
1
4 w™
1
10 5  14= 10
1

2 w¡
5
2
3
4 b™
3
10 5  34= 10
3

S 3
2
4 w™
3
10 5  24= 10
3
3
5 b¡
3
2
4 b™
3
10 5  24= 10
3

Figure 7 Sum= 1

*The sample space for the combined outcomes is S = 5w1w2, w1b2, b1w2, b1b2 6. If we let
w1 = 5w1w2, w1b2 6 and w2 = 5w1w2, b1w2 6, then w1 ¨ w2 = 5w1w2 6.
432 CHAPTER 8 Probability

Now we can complete the problem. A white ball drawn on the second draw
corresponds to either the combined outcome w1 ¨ w2 or b1 ¨ w2 occurring. Thus,
since these combined outcomes are mutually exclusive,
P1w2 2 = P1w1 ¨ w2 2 + P1b1 ¨ w2 2
1 3 4 2
= + = =
10 10 10 5
which is the sum of the probabilities listed at the ends of the two paths terminating
in w2.

Matched Problem 4 Two balls are drawn in succession without replacement from
a box containing 4 red and 2 white balls. What is the probability of drawing a red
ball on the second draw?

The sequence of two experiments in Example 4 is an example of a stochastic

process. In general, a stochastic process involves a sequence of experiments where
the outcome of each experiment is not certain. Our interest is in making predictions
about the process as a whole. The analysis in Example 4 generalizes to stochastic
processes involving any finite sequence of experiments. We summarize the proce-
dures used in Example 4 for general application:

ProceDure Constructing Probability Trees

Step 1 Draw a tree diagram corresponding to all combined outcomes of the
sequence of experiments.
Step 2 Assign a probability to each tree branch. (This is the probability of the occur-
rence of the event on the right end of the branch subject to the occurrence of
all events on the path leading to the event on the right end of the branch. The
probability of the occurrence of a combined outcome that corresponds to a
path through the tree is the product of all branch probabilities on the path.*)
Step 3 Use the results in Steps 1 and 2 to answer various questions related to the
sequence of experiments as a whole.
*If we form a sample space S such that each simple event in S corresponds to one path through the tree,
and if the probability assigned to each simple event in S is the product of the branch probabilities on the
corresponding path, then it can be shown that this is not only an acceptable assignment (all probabilities
for the simple events in S are nonnegative and their sum is 1), but it is the only assignment consistent
with the method used to assign branch probabilities within the tree.

Explore and Discuss 1 Refer to the table on rain and accidents in Example 2 and use formula (1), where
appropriate, to complete the following probability tree:

A .025 ? A .025 RA
.360 R .360 R
A' ? A' ? RA'
S S
? A ? R'A
A ?
R' R'
? A' ? R'A'
A'

Discuss the difference between P1R ¨ A2 and P1A∙ R2.

 SECTION 8.3 Conditional Probability, Intersection, and Independence 433

ExamplE 5 Product Defects An auto company A subcontracts the manufac-

turing of its onboard computers to two companies: 40% to company B and 60%
to company C. Company B in turn subcontracts 70% of the orders it receives from
company A to company D and the remaining 30% to company E, both subsidiar-
ies of company B. When the onboard computers are completed by companies D, E,
and C, they are shipped to company A to be used in various car models. It has been
found that 1.5%, 1%, and .5% of the boards from D, E, and C, respectively, prove
defective during the 3-year warranty period after a car is first sold. What is the
probability that a given onboard computer will be defective during the 3-year war-
ranty period?
Solution Draw a tree diagram and assign probabilities to each branch (Fig. 8):

.985 OK
D
.7
.015 Defective
B
.4 .99 OK
.3
E
A .01 Defective

.6 .995 OK
C
Figure 8 .005 Defective

There are three paths leading to defective (the onboard computer will be
defective within the 3-year warranty period). We multiply the branch probabilities
on each path and add the three products:
P1defective2 = 1.421.721.0152 + 1.421.321.012 + 1.621.0052
= .0084

Matched Problem 5 In Example 5, what is the probability that a given onboard

computer came from company E or C?

Independent Events
We return to Example 4, which involved drawing two balls in succession without
replacement from a box of 3 blue and 2 white balls. What difference does “without
replacement” and “with replacement” make? Figure 9 shows probability trees cor-
responding to each case. Go over the probability assignments for the branches in
Figure 9B to convince yourself of their correctness.

FIRST SECOND COMBINED SECOND FIRST

DRAW DRAW OUTCOMES DRAW DRAW
1 2
4 w™ .10 w¡w™ .16 w™ 5

2 w¡ w¡ 2
5 3 3 5
4 b™ .30 w¡b™ .24 b™ 5
S 2 2 S
4 w™ .30 b¡w™ .24 w™ 5
3 3
5 b¡ b¡ 5

2 3
4 b™ .30 b¡b™ .36 b™ 5

(A) Without replacement (B) With replacement

Figure 9 S = 5 W1W2, W1b2, b1W2, b1b2 6

434 CHAPTER 8 Probability

Let
A = white ball on second draw = 5w1w2, b1w2 6
B = white ball on first draw = 5w1w2, w1b2 6
We now compute P1A∙ B2 and P(A) for each case in Figure 9.

Case 1. Without replacement:

P1A ¨ B2 P5w1w2 6 .10
P1A ∙ B2 = = = = .25
P1B2 P5w1w2, w1b2 6 .10 + .30
(This is the assignment to branch w1w2 that we made by looking in the box and
counting.)
P1A2 = P5w1w2, b1w2 6 = .10 + .30 = .40
Note that P1A∙ B2 ≠ P1A2, and we conclude that the probability of A is affected by
the occurrence of B.

Case 2. With replacement:

P1A ¨ B2 P5w1w2 6 .16
P1A ∙ B2 = = = = .40
P1B2 P5w1w2, w1b2 6 .16 + .24
(This is the assignment to branch w1w2 that we made by looking in the box and
counting.)
P1A2 = P5w1w2, b1w2 6 = .16 + .24 = .40
Note that P1A∙ B2 = P1A2, and we conclude that the probability of A is not affected
by the occurrence of B.
Intuitively, if P1A∙ B2 = P1A2, then it appears that event A is “independent” of
B. Let us pursue this further. If events A and B are such that
P1A∙ B2 = P1A2
then replacing the left side by its equivalent from formula (1), we obtain
P1A ¨ B2
= P1A2
P1B2
After multiplying both sides by P(B), the last equation becomes
P1A ¨ B2 = P1A2P1B2
This result motivates the following definition of independence:

Definition Independence
If A and B are any events in a sample space S, we say that A and B are indepen-
dent if
P1A ¨ B2 = P1A2P1B2 (3)
Otherwise, A and B are said to be dependent.

From the definition of independence one can prove (see Problems 75 and 76,
Exercises 8.3) the following theorem:

theorem 2 On Independence
If A and B are independent events with nonzero probabilities in a sample space S,
then
P1A∙ B2 = P1A2 and P1B∙ A2 = P1B2 (4)
If either equation in (4) holds, then A and B are independent.
 SECTION 8.3 Conditional Probability, Intersection, and Independence 435

ConCEptual i n S i g h t
Sometimes intuitive reasoning can be helpful in deciding whether or not two events are
independent. Suppose that a fair coin is tossed five times. What is the probability of a
head on the fifth toss, given that the first four tosses are all heads? Our intuition tells us
that a coin has no memory, so the probability of a head on the fifth toss given four pre-
vious heads should be equal to the probability of a head on the fifth toss, namely, 1/2.
In other words, the first equation of Theorem 2 holds intuitively, so “heads on the fifth
toss” and “heads on the first four tosses” are independent events.
Often, unfortunately, intuition is not a reliable guide to the notion of independence.
Independence is a technical concept. So in all cases, an appropriate sample space should
be chosen, and either equation (3) or equation (4) should be tested, to confirm that two
events are (or are not) independent.

ExamplE 6 Testing for Independence In two tosses of a single fair coin,

show that the events “A head on the first toss” and “A head on the second toss” are
independent.
Solution Consider the sample space of equally likely outcomes for the tossing
of a fair coin twice,
S = 5HH, HT, TH, TT6
and the two events,
A = a head on the first toss = 5HH, HT6
B = a head on the second toss = 5HH, TH6
Then
2 1 2 1 1
P1A2 = = P1B2 = = P1A ¨ B2 =
4 2 4 2 4
Thus,

= # = P1A2P1B2
1 1 1
P1A ¨ B2 =
4 2 2
and the two events are independent. (The theory agrees with our intuition—a coin
has no memory.)

Matched Problem 6 In Example 6, compute P1B∙ A2 and compare with P(B).

ExamplE 7 Testing for Independence A single card is drawn from a standard

52-card deck. Test the following events for independence (try guessing the answer
to each part before looking at the solution):
(A) E = the drawn card is a spade.
F = the drawn card is a face card.
(B) G = the drawn card is a club.
H = the drawn card is a heart.
Solution
(A) To test E and F for independence, we compute P1E ¨ F2 and P(E)P(F).
If they are equal, then events E and F are independent; if they are not equal,
then events E and F are dependent.
3 13 12 3
P1E ¨ F2 = P1E2P1F2 = a ba b =
52 52 52 52
Events E and F are independent. (Did you guess this?)
436 CHAPTER 8 Probability

(B) Proceeding as in part (A), we see that

13 13 1
P1G ¨ H2 = P1∅2 = 0 P1G2P1H2 = a ba b =
52 52 16
Events G and H are dependent. (Did you guess this?)

! caution Students often confuse mutually exclusive (disjoint) events with in-
dependent events. One does not necessarily imply the other. In fact, it is not difficult
to show (see Problem 79, Exercises 8.3) that any two mutually exclusive events A
and B, with nonzero probabilities, are always dependent. ▲

Matched Problem 7 A single card is drawn from a standard 52-card deck. Test
the following events for independence:
(A) E = the drawn card is a red card
F = the drawn card=s number is divisible by 5 1face cards are not assigned values2
(B) G = the drawn card is a king
H = the drawn card is a queen

Explore and Discuss 2 In college basketball, would it be reasonable to assume that the following events are
independent? Explain why or why not.
A = the Golden Eagles win in the first round of the NCAA
tournament.
B = the Golden Eagles win in the second round of the NCAA
tournament.

The notion of independence can be extended to more than two events:

Definition Independent Set of Events

A set of events is said to be independent if for each finite subset 5E1, E2, c, Ek 6
P1E1 ¨ E2 ¨ g ¨ Ek 2 = P1E1 2P1E2 2 # g # P1Ek 2 (5)
The next example makes direct use of this definition.

ExamplE 8 Computer Control Systems A space shuttle has four independent

computer control systems. If the probability of failure (during flight) of any one
system is .001, what is the probability of failure of all four systems?
Solution Let

E1 = failure of system 1 E3 = failure of system 3

E2 = failure of system 2 E4 = failure of system 4
Then, since events E1, E2, E3, and E4 are given to be independent,
P1E1 ¨ E2 ¨ E3 ¨ E4 2 = P1E1 2P1E2 2P1E3 2P1E4 2
= 1.0012 4
= .000 000 000 001

Matched Problem 8 A single die is rolled 6 times. What is the probability of

getting the sequence 1, 2, 3, 4, 5, 6?
 SECTION 8.3 Conditional Probability, Intersection, and Independence 437

Summary
The key results in this section are summarized in the following box:

Summary Key Concepts

Conditional Probability
P1A ¨ B2 P1B ¨ A2
P1A∙ B2 = P1B∙ A2 =
P1B2 P1A2
Note: P1A 0 B2 is a probability based on the new sample space B, while P1A ¨ B2
and P(B) are probabilities based on the original sample space S.
Product Rule
P1A ¨ B2 = P1A2P1B∙ A2 = P1B2P1A∙ B2
Independent Events
• A and B are independent if
P1A ¨ B2 = P1A2P1B2

• If A and B are independent events with nonzero probabilities, then

P1A∙ B2 = P1A2 and P1B∙ A2 = P1B2

• If A and B are events with nonzero probabilities and either P1A∙ B2 = P1A2
or P1B∙ A2 = P1B2, then A and B are independent.
• If E1, E2, c, En are independent, then
P1E1 ¨ E2 ¨ g ¨ En 2 = P1E1 2P1E2 2 # g # P1En 2

Exercises 8.3
Skills Warm-up Exercises 10. The card is a face card, given that it is red.
W In Problems 1–6, use a tree diagram to represent a factorization 11. The card is black, given that it is a club.
of the given integer into primes, so that there are two branches
12. The card is a jack, given that it is red.
at each number that is not prime. For example, the factorization
24 = 4 # 6 = 12 # 22 # 12 # 32 is represented by: 13. The card is a club, given that it is black.
2 14. The card is red, given that it is a jack.
4
2
24
2
6 In Problems 15–22, find the conditional probability, in a single
3
roll of two fair dice, that
(If necessary, review Section A.3.)
15. The sum is less than 6, given that the sum is even.
1. 100 2. 120
16. The sum is 10, given that the roll is doubles.
3. 180 4. 225
17. The sum is even, given that the sum is less than 6.
5. 315 6. 360
18. The roll is doubles, given that the sum is 10.
A single card is drawn from a standard 52-card deck. In Problems 19. The sum is greater than 7, given that neither die is a
7–14, find the conditional probability that six.
7. The card is an ace, given that it is a heart. 20. The sum is odd, given that at least one die is a six.
8. The card is red, given that it is a face card. 21. Neither die is a six, given that the sum is greater than 7.
9. The card is a heart, given that it is an ace. 22. At least one die is a six, given that the sum is odd.
438 CHAPTER 8 Probability

In Problems 23–42, use the table below. Events A, B, and C are (A) Find P1F ∙ E2.
mutually exclusive; so are D, E, and F.
A B C Totals
(B) Test events E and F for independence.
D .20 .03 .07 .30 46. Repeat Problem 45 with the following events:
E .28 .05 .07 .40 E = pointer lands on an odd number
F .22 .02 .06 .30 F = pointer lands on a prime number
Totals .70 .10 .20 1.00
Compute the indicated probabilities in Problems 47 and 48 by
In Problems 23–26, find each probability directly from the table. referring to the following probability tree:
.4 R
23. P1B2 24. P1E2
.3 M
25. P1B ¨ D2 26. P1C ¨ E2 .6 S
.2 R
In Problems 27–34, compute each probability using formula (1) .7 N
on page 428 and appropriate table values. .8 S
27. P1D∙ B2 28. P1C∙ E2 47. (A) P1M ¨ S2 (B) P1R2
29. P1B ∙ D2 30. P1E ∙ C2 48. (A) P1N ¨ R2 (B) P1S2
31. P1D∙ C2 32. P1E ∙ A2 49. A fair coin is tossed twice. Consider the sample space
33. P1A ∙ C2 34. P1B ∙ B2 S = 5HH, HT, TH, TT6 of equally likely simple events.
We are interested in the following events:
In Problems 35–42, test each pair of events for independence. E1 = a head on the first toss
35. A and D 36. A and E E2 = a tail on the first toss
37. B and D 38. B and E E3 = a tail on the second toss
39. B and F 40. C and F E4 = a head on the second toss
41. A and B 42. D and F For each pair of events, discuss whether they are independent
and whether they are mutually exclusive.
43. A fair coin is tossed 8 times.
(A) E1 and E4
(A) What is the probability of tossing a head on the 8th toss,
given that the preceding 7 tosses were heads? (B) E1 and E2
(B) What is the probability of getting 8 heads or 8 tails? 50. For each pair of events (see Problem 49), discuss whether
they are independent and whether they are mutually exclusive.
44. A fair die is rolled 5 times.
(A) E1 and E3
(A) What is the probability of getting a 6 on the 5th roll,
given that a 6 turned up on the preceding 4 rolls? (B) E3 and E4
(B) What is the probability that the same number turns up 51. In 2 throws of a fair die, what is the probability that you will
every time? get an even number on each throw? An even number on the
first or second throw?
45. A pointer is spun once on the circular spinner shown below. The
probability assigned to the pointer landing on a given integer 52. In 2 throws of a fair die, what is the probability that you will get
(from 1 to 5) is the ratio of the area of the corresponding circular at least 5 on each throw? At least 5 on the first or second throw?
sector to the area of the whole circle, as given in the table:
53. Two cards are drawn in succession from a standard 52-card
ei 1 2 3 4 5 deck. What is the probability that the first card is a club and
P(ei) .3 .1 .2 .3 .1 the second card is a heart
(A) If the cards are drawn without replacement?
(B) If the cards are drawn with replacement?
54. Two cards are drawn in succession from a standard 52-card
deck. What is the probability that both cards are red
(A) If the cards are drawn without replacement?
(B) If the cards are drawn with replacement?
55. A card is drawn at random from a standard 52-card deck.
Events G and H are
Given the events G = the drawn card is black.
E = pointer lands on an even number H = the drawn card is divisible by 3
F = pointer lands on a number less than 4 1face cards are not valued2.
 SECTION 8.3 Conditional Probability, Intersection, and Independence 439

(A) Find P1H∙ G2. 65. If A is nonempty and A ⊂ B, then P1A ∙ B2 Ú P1A2.
(B) Test H and G for independence. 66. If A and B are events, then P1A ∙ B2 … P1B2.
56. A card is drawn at random from a standard 52-card deck. 67. If A and B are mutually exclusive, then A and B are
Events M and N are independent.
M = the drawn card is a diamond. 68. If A and B are independent, then A and B are mutually
N = the drawn card is even exclusive.
1face cards are not valued2. 69. If two balls are drawn in succession, with replacement, from a
(A) Find P1N ∙ M2. box containing m red and n white balls 1m Ú 1 and n Ú 12,
then
(B) Test M and N for independence.
P1W1 ¨ R2 2 = P1R1 ¨ W2 2
57. Let A be the event that all of a family’s children are the same
gender, and let B be the event that the family has at most 1 70. If two balls are drawn in succession, without replacement, from
boy. Assuming the probability of having a girl is the same as a box containing m red and n white balls 1m Ú 1 and n Ú 12,
the probability of having a boy (both .5), test events A and B then
for independence if
P1W1 ¨ R2 2 = P1R1 ¨ W2 2
(A) The family has 2 children.
(B) The family has 3 children. 71. A box contains 2 red, 3 white, and 4 green balls. Two balls
are drawn out of the box in succession without replacement.
58. An experiment consists of tossing n coins. Let A be the event What is the probability that both balls are the same color?
that at least 2 heads turn up, and let B be the event that all the
coins turn up the same. Test A and B for independence if 72. For the experiment in Problem 71, what is the probability that
no white balls are drawn?
(A) 2 coins are tossed.
(B) 3 coins are tossed. 73. An urn contains 2 one-dollar bills, 1 five-dollar bill, and 1
ten-dollar bill. A player draws bills one at a time without
Problems 59–62 refer to the following experiment: 2 balls are replacement from the urn until a ten-dollar bill is drawn.
drawn in succession out of a box containing 2 red and 5 white Then the game stops. All bills are kept by the player.
balls. Let Ri be the event that the ith ball is red, and let Wi be the
event that the ith ball is white. (A) What is the probability of winning $16?

59. Construct a probability tree for this experiment and find the (B) What is the probability of winning all bills in the urn?
probability of each of the events R1 ¨ R2, R1 ¨ W2, W1 ¨ R2, (C) What is the probability of the game stopping at the
W1 ¨ W2, given that the first ball drawn was second draw?
(A) Replaced before the second draw 74. Ann and Barbara are playing a tennis match. The first
(B) Not replaced before the second draw player to win 2 sets wins the match. For any given set, the
probability that Ann wins that set is 23. Find the probability
60. Find the probability that the second ball was red, given that that
the first ball was
(A) Ann wins the match.
(A) Replaced before the second draw
(B) 3 sets are played.
(B) Not replaced before the second draw
(C) The player who wins the first set goes on to win the
61. Find the probability that at least 1 ball was red, given that the match.
first ball was
75. Show that if A and B are independent events with nonzero
(A) Replaced before the second draw probabilities in a sample space S, then
(B) Not replaced before the second draw P1A ∙ B2 = P1A2 and P1B ∙ A2 = P1B2
62. Find the probability that both balls were the same color,
76. Show that if A and B are events with nonzero probabili-
given that the first ball was
ties in a sample space S, and either P1A ∙ B2 = P1A2 or
(A) Replaced before the second draw P1B ∙ A2 = P1B2, then events A and B are independent.
(B) Not replaced before the second draw 77. Show that P1A ∙ A2 = 1 when P1A2 ≠ 0.
In Problems 63–70, discuss the validity of each statement. If the 78. Show that P1A ∙ B2 + P1A′∙ B2 = 1.
statement is always true, explain why. If not, give a counterexample. 79. Show that A and B are dependent if A and B are mutually
63. If P1A ∙ B2 = P1B2, then A and B are independent. exclusive and P1A2 ≠ 0, P1B2 ≠ 0.
64. If A and B are independent, then P1A ∙ B2 = P1B ∙ A2. 80. Show that P1A ∙ B2 = 1 if B is a subset of A and P1B2 ≠ 0.