Discrete random variables

DISCRETE PROBABILITY

1. Let X be a discrete random variable with the following probability

distribution. Find P(X ≥ 3).

X 1 2 3 4 5

P(X) 0.35 0.25 0.20 0.15 ?

Solution:

First, we need to find the P(X = 5), which we’ll do by subtracting all the
other probabilities from 1.

P(X = 5) = 1 − 0.35 − 0.25 − 0.20 − 0.15

P(X = 5) = 1 − 0.95

P(X = 5) = 0.05

Then the probability that X ≥ 3 is

P(X ≥ 3) = P(X = 3) + P(X = 4) + P(X = 5)

P(X ≥ 3) = 0.20 + 0.15 + 0.05

P(X ≥ 3) = 0.40

1
 2. Let B be a discrete random variable with the following probability
distribution. Find μB and σB.

B 0 5 10 15

P(B) 1/5 1/5 2/5 1/5

Solution:

We’ll weight each value of B by the probability that the value occurs, P(B),
in order to find the expected value μB.

(5) (5) (5) (5)

4
1 1 2 1
∑ i
μB = E(B) = B P(Bi) = 0 +5 + 10 + 15
i=1

μB = 8

In order to find the standard deviation of B, σB, we have to find variance

first.
4
σB2 = (Bi − μB)2 P(Bi)
∑
i=1

(5) (5) (5) (5)

1 2 1 2 2 2 1
σB2 = (0 − 8) 2
+ (5 − 8) + (10 − 8) + (15 − 8)

σB2 = 26

Then the standard deviation is

2
 σB2 = 26

σB ≈ 5.099

3. The table shows the distribution of size of households in the U.S. for
2016. Suppose we select a household of size at least 2 at random. What is
the probability that this household has a size of at least 4?

Size of 1 2 3 4 5 6 7+
household
P(size) 0.28 0.34 ? 0.13 0.06 0.02 0.01

Solution:

Find the probability that the household is a 3-person household.

P(X = 3) = 1 − 0.281 − 0.340 − 0.129 − 0.060 − 0.023 − 0.013

P(X = 3) = 0.154

The probability of “at least 4” is

P(at least 4) = P(X = 4) + P(X = 5) + P(X = 6) + P(X = 7)

P(at least 4) = 0.129 + 0.060 + 0.023 + 0.013

P(at least 4) = 0.225

and the probability of “at least 2” is

3
 P = (at least 2) = 1 − P(X = 1)

P = (at least 2) = 1 − 0.281

P = (at least 2) = 0.719

Then the probability that the household size is at least 4, given that the
household size is at least 2, is

P(size at least 4)
P(at least 4, given at least 2) =
P(size at least 2)

0.225
P(at least 4, given at least 2) = ≈ 0.313
0.719

4. A standard deck of cards is shuffled, and two cards are selected

without replacement. Let R be the number of red cards selected. Construct
a probability distribution for R.

Solution:

If we draw two cards, we can find the probability that either both are red
P(R = 2), that one is red P(R = 1), or that neither are red P(R = 0).

52 ( 51 ) 102
26 25 25
P(R = 0) = =

52 ( 51 ) 102
26 25 25
P(R = 2) = =

4
Then the probability that one card is red is

P(R = 1) = 1 − P(R = 2) − P(R = 0)

25 25
P(R = 1) = 1 − −
102 102

102 25 25
P(R = 1) = − −
102 102 102

52 26
P(R = 1) = =
102 51

Which means we can build a probability distribution for R.

R 0 1 2

P(R) 25/102 52/102 25/102

5. A local restaurant features a wheel we can spin before paying the bill.
The wheel is split into 8 equal size pieces. One of the sections gives us a
$10 discount on the bill, two sections give a $5 discount, three sections give
a $2 discount, and the rest of the sections give no discount. Find the
expected value for the discount given by the wheel.

Solution:

Let X be the amount of the discount. Then the expected value, or mean of
the discount is

5
 (8) (8) (8) (8)
1 2 3 2
∑
E(X ) = XP(X ) = 10 +5 +2 +0

E(X ) = $3.25

6. John stops at the local gas station and decides to buy lottery tickets.
Each ticket has a 20 % chance of being a winner. He will buy a lottery ticket
and check to see if it’s a winner. If it’s a winner, he’ll collect his money and
be done. If it’s not a winner, he’ll buy another. He’ll repeat this until he gets
a winning ticket. But if he hasn’t won by his fifth ticket, he won’t buy any
more tickets. Let L be the number of lottery tickets John will buy, then find
E(L).

Solution:

We could find the probability of winning on each of the first four tickets.

P(L = 1) = 0.2

P(L = 2) = (0.2)(0.8) = 0.16

P(L = 3) = (0.2)(0.8)(0.8) = 0.128

P(L = 4) = (0.2)(0.8)(0.8)(0.8) = 0.1024

If we continue this pattern, we might think that the probability of winning

on the fifth ticket would be

P(L = 5) = (0.2)(0.8)(0.8)(0.8)(0.8) = 0.08192

6
But the question tells us that John will never buy more than five tickets.
Because he’s guaranteed to buy one, two, three, four, or five tickets, the
probability that he’s going to purchase one of those numbers of tickets
must be 100 % . So the probability that he purchases five tickets is actually

P(L = 5) = 1 − P(L ≤ 4)

P(L = 5) = 1 − (0.2 + 0.16 + 0.128 + 0.1024)

P(L = 5) = 1 − 0.5904

P(L = 5) = 0.4096

Then the expected value for the number of tickets he’ll buy, L, is

E(L) = 1(0.2) + 2(0.16) + 3(0.128) + 4(0.1024) + 5(0.4096)

E(L) = 0.2 + 0.32 + 0.384 + 0.4096 + 2.048

E(L) = 3.3616

7
TRANSFORMING RANDOM VARIABLES

1. We use the formula

∘ 9∘
F= C + 32
5

to convert from Celsius to Fahrenheit. August is the hottest month in

Hawaii with a mean temperature of 27∘C. What is the mean temperature in
Hawaii in ∘F.

Solution:

We’ll plug 27∘C into the conversion formula to get the corresponding value
in Fahrenheit.

9 9 243
μ∘F = μ∘C + 32 = (27) + 32 = + 32 = 80.6∘
5 5 5

2. Let Z be a random variable with σZ2 = 49. Let W = (1/2)Z − 10. Find σW.

Solution:

We’ve been given the variance of Z, so we need to use it first to find the
standard deviation of Z.

8
 σZ2 = 49

σZ = 7

Standard deviation is effected by scaling, but not by shifting. So when we

convert from Z to W using

1
W= Z − 10
2

we need to multiply by 1/2, but we don’t need to shift by 10. So we can say
that the standard deviation σW is

1
σW = σZ
2

1
σW = (7)
2

7
σW =
2

3. The students in each 8th period classroom were asked to donate

money for a school fundraiser, and the class that raised the most money
was awarded a pizza party. The school secretary recorded the amount
raised by each class and made a five-number summary for the data.

Min Q1 Median Q3 Max

4.50 15.25 22.00 38.75 95.50

9
If a donor commits to matching equally the students’ donations, create a
new five-number summary of the total amount raised, including the
donor’s contribution.

Solution:

The donor is essentially scaling the data set, because she’s doubling the
students’ donations. And we know that scaling a data set scales all the
values in the five-number summary.

Therefore, after putting the donor’s contribution together with the

students’ donations, we can give a the new five-number summary for the
data set as

Min Q1 Median Q3 Max

4.5(2) 15.25(2) 22(2) 38.75(2) 95.5(2)

Min Q1 Median Q3 Max

9 30.5 44 77.5 191

4. The number of items sold at a concession stand is normally

distributed with μ = 323 and σ = 30. The average price per item sold is $1.25.
Different student clubs volunteer to work the concession stand throughout
the year and get to keep half of their sales to go towards their club’s
activities. What is the probability that a club will get to keep more than
$220 in sales?

10
Solution:

Let N be the number of items sold and S be the amount of money the club
gets to keep. Then we could write an equation for the amount of money
they keep as

1
S= (1.25)(N )
2

We know the mean of N is μN = 323, so we can use the conversion equation

to find the mean of S.

1 1
μS = (1.25)(μN ) = (1.25)(323) ≈ 201.88
2 2

Using σN = 30, we can find the standard deviation of S in the same way.

1 1
σS = (1.25)(σN ) = (1.25)(30) = 18.75
2 2

With a mean of μS = 201.88 and a standard deviation of σS = 18.75, we can

find the probability that the club will take home more than $220. We’ll find
the z-score associated with $220.

220 − 201.88
z= = 0.9664 ≈ 0.97
18.75

Since we’re looking at 0.97 standard deviations above the mean, we get
0.8340 from the z-table, which tells us that the probability that the club
takes home more than $220 is 1 − 0.8340 = 0.166 ≈ 17 % . They have an
approximately 17 % chance of taking home more than $220.

11
 5. The average length of a full-term new born baby is 20 inches with
variance 0.81 inches. What are the mean and standard deviation of the
length of a full-term new born, expressed in centimeters? Use
1 in = 2.54 cm.

Solution:

To convert between inches and centimeters, we’ll say

μlength in cm = 2.54μlength in in

μlength in cm = 2.54(20)

μlength in cm = 50.8 cm

We’ll use the same conversion formula to convert the standard deviation.

σlength in cm = 2.54σlength in in

σlength in cm = 2.54 0.81

σlength in cm = 2.286 cm

6. The weights of full-term new born babies are normally distributed with
μ = 120 ounces and σ = 20 ounces. Describe the shape, center, and spread

12
for the weights of full-term new born babies as measured in pounds. Use
1 pound = 16 ounces.

Solution:

We can use a conversion formula to convert the mean from pounds to

ounces.

1
μweight in pounds = μ
16 weight in ounces

1
μweight in pounds = (120)
16

μweight in pounds = 7.5 pounds

Now we’ll convert the given standard deviation.

1
σweight in pounds = σ
16 weight in ounces

1
σweight in pounds = (20)
16

σweight in pounds = 1.25 pounds

The distribution of weights of full-term new born babies remains normally

distributed, even after converting from ounces to pounds. The mean is
μ = 7.5 pounds and the standard deviation is σ = 1.25 pounds.

13
COMBINATIONS OF RANDOM VARIABLES

1. X and Y are independent random variables with E(X ) = 48, E(Y ) = 54,
SD(X ) = 3 and SD(Y ) = 5. Find E(X − Y ) and SD(X − Y ).

Solution:

To find the expected value of the difference, we find the difference of the
expected values.

E(X − Y ) = E(X ) − E(Y ) = 48 − 54 = − 6

To find the standard deviation of the difference, we have to square both

standard deviations in order to get the variances. We get SD 2(X ) = 32 = 9
and SD 2(Y ) = 52 = 25. Then we can find the standard deviation of the
difference.

SD(X − Y ) = SD 2(X ) + SD 2(Y ) = 32 + 52 = 34 ≈ 5.831

2. A and B are independent random variables with E(A) = 6.5, E(B) = 4.4,
SD(A) = 1.6, and SD(B) = 2.1. Find E(4A + 2B) and SD(4A + 2B).

Solution:

14
The expected value of the sum of variables is the sum of the expected
values.

E(4A + 2B) = 4E(A) + 2E(B) = 4(6.5) + 2(4.4) = 34.8

Then we’ll find the standard deviation of the combination. When we scale
a variable by some constant k, its standard deviation gets scaled by k as
well. So the standard deviations of 4A and 2B are

SD(A) = 1.6

SD(4A) = 4(1.6)

SD(4A) = 6.4

and

SD(B) = 2.1

SD(2B) = 2(2.1)

SD(2B) = 4.2

Then the variances of the variables 4A and 2B are

SD 2(4A) = 6.42

SD 2(4A) = 40.96

and

SD 2(2B) = 4.22

SD 2(2B) = 17.64

15
The variance of a combination is the sum of the variances, so

SD 2(4A) + SD 2(2B) = 40.96 + 17.64

SD 2(4A) + SD 2(2B) = 58.6

Then the standard deviation of the combination is

SD 2(4A) + SD 2(2B) = 58.6

SD 2(4A) + SD 2(2B) ≈ 7.66

3. The time it takes students to complete multiple choice questions on an

AP Statistics Exam has a mean of 55 seconds with a standard deviation of
12 seconds. If the exam consists of 40 multiple choice questions, find the
mean total time to finish the exam. Then find the standard deviation in the
total time. What assumption must be made?

Solution:

We have to assume that the questions are independent. Then we can say
that the mean finishing time is

μQ1 + μQ2 + μQ3 + . . . + μQ40 = 40(55) = 2,200 ≈ 36.67 minutes

and that the variance of the finishing time is

σQ2 1 + σQ2 2 + σQ2 3 + . . . + σQ2 40 = 40(122) = 40(144) = 5,760 seconds

16
such that the standard deviation of the finishing time is

σ= 5,760 ≈ 75.89 ≈ 1.26 minutes

4. Let M represent the height of a male over 21 years of age and let W
represent the height of a female over 21 years of age. Let D represent the
difference between their heights (D = M − W). Let E(M ) = 70 inches, σM = 2.8
inches, E(W ) = 64.5 inches and σW = 2.4 inches.

What is the mean and standard deviation of the difference between the
two heights?

Solution:

To find the mean of the difference, we’ll find the difference of the means.

E(M − W ) = E(M ) − E(W ) = 70 − 64.5 = 5.5 inches

We’ll find variance in order to get standard deviation. The variances are
σM2 = 2.82 = 7.84 and σW2 = 2.42 = 5.76. Therefore, the standard deviation of
the difference is

σ(M − W ) = σM2 + σW2 = 7.84 + 5.76 = 13.6 ≈ 3.69 inches

5. The Ironman is a challenge in which a competitor swims 2.4 miles,

then bikes 112 miles, and finally runs 26.2 miles. Suppose the times for each

17
of the legs are normally distributed with the given means and standard
deviations.

Swim: μS = 76 minutes and σS = 18 minutes

Bike: μB = 385 minutes and σB = 32 minutes

Run: μR = 294 minutes and σR = 25 minutes

What percent of the competitors finish the Ironman in under 710 minutes?

Solution:

Let T be the total time to complete all three legs of the Ironman. Then the
mean finishing time is

μT = μS + μB + μR = 76 + 385 + 294 = 755 minutes

Assuming the legs are independent random variables, then we can find the
sum of the variances to get the variance of the sum.

σT2 = σS2 + σB2 + σR2 = 182 + 322 + 252 = 1,973

Then the standard deviation of finishing time is

σT = 1,973 ≈ 44.42 minutes

Since S, B, and R are normally distributed, T will also be normally

distributed. To find the probability that a finisher will finish in under 710
minutes, we’ll find the z-score associated with 710 minutes.

18
 710 − 755
z= ≈ − 1.01
44.42

If we look up a z-score of z = − 1.01 in a z-table, we get 0.1562, which means

there’s an approximately 15.62 % chance that a finisher finishes in under
710 minutes.

6. We buy a scratch-off lottery ticket for $1 at the local gas station. If we

get three hearts in a row on the scratch-off, the state will pay us $500. Let
X be the amount the state pays us and let X have the following probability
distribution.

X $0 $500

P(X) 0.999 0.001

Suppose we buy one of these scratch-off tickets every day for a week (7
days). Find the expected value and standard deviation of our total
winnings.

Solution:

The expected value of our winnings on any one ticket is

E(X ) = 0(0.999) + 500(0.001) = $0.50

Find the standard deviation of the winnings by taking the sum of the
variances, weighted by the associated probabilities.

19
 SD(X ) = (0 − 0.5)2(0.999) + (500 − 0.5)2(0.001) = 249.75 ≈ 15.80

Let W be the amount the state pays us for 7 lottery tickets. The expected
value of the total winnings for 7 lottery tickets is therefore
E(W ) = 7(0.5) = $3.50. The standard deviation of the total winnings is

SD(W ) = (15.80)2 + (15.80)2 + . . . + (15.80)2

SD(W ) = 7(15.80)2

SD(W ) = 1,747.48

SD(W ) ≈ $41.80

These are the mean and standard deviation of total winnings. They don’t
account for the price we paid for the scratch-off tickets. If we want to
account for the cost of the tickets in order to calculate profit, instead of
just winnings, then we’d use the probability distribution

X -$1 $499

P(X) 0.999 0.001

in order to calculate the mean and standard deviation of our profit.

20
PERMUTATIONS AND COMBINATIONS

1. Calculate the binomial coefficient.

(7)
12

Solution:

Use the combination formula

(k)
n n!
= nCk =
k!(n − k)!

Plug in n = 12 and k = 7.

( 7 ) 12 7 7! ⋅ 5!
12 12! 12 ⋅ 11 ⋅ 10 ⋅ 9 ⋅ 8 ⋅ 7 ⋅ 6 ⋅ 5 ⋅ 4 ⋅ 3 ⋅ 2 ⋅ 1
= C = =
7⋅6⋅5⋅4⋅3⋅2⋅1⋅5⋅4⋅3⋅2⋅1

(7)
12 12 ⋅ 11 ⋅ 10 ⋅ 9 ⋅ 8
=
5⋅4⋅3⋅2

(7)
12
= 792

2. Calculate 10 P3.

21
Solution:

Use the permutation formula

n!
n Pk =
(n − k)!

Plug in n = 10 and k = 3.

10! 10!
10 P3 = =
(10 − 3)! 7!

10 ⋅ 9 ⋅ 8 ⋅ 7 ⋅ 6 ⋅ 5 ⋅ 4 ⋅ 3 ⋅ 2 ⋅ 1
10 P3 =
7⋅6⋅5⋅4⋅3⋅2⋅1

10 P3 = 10 ⋅ 9 ⋅ 8

10 P3 = 720

3. How much greater is 5P2 than 5C2?

Solution:

We’ll calculate both values, then find the difference.

5! 5! 5 ⋅ 4 ⋅ 3 ⋅ 2 ⋅ 1
5P2 = = = = 5 ⋅ 4 = 20
(5 − 2)! 3! 3⋅2⋅1

5! 5! 5⋅4⋅3⋅2⋅1 5⋅4
5C2 = = = = = 10
2!(5 − 2)! 2! ⋅ 3! 2 ⋅ 1 ⋅ 3 ⋅ 2 ⋅ 1 2

22
The difference between 5P2 and 5C2 is

5P2 −5 C2 = 20 − 10 = 10

4. The high school girls’ basketball team has 8 players, 5 of whom are
seniors. They need to figure out which senior will be captain and which
senior will be co-captain. To make it fair, they choose two players out of a
hat. The first drawn will be captain and the second will be co-captain. How
many different captain/co-captain pairs are possible?

Solution:

Since the order matters, we have to calculate the permutations. There are
5 seniors we can choose from, and 2 spots to put them in.

n! 5! 5! 5 ⋅ 4 ⋅ 3 ⋅ 2 ⋅ 1
n Pk = = = = = 5 ⋅ 4 = 20
(n − k)! (5 − 2)! 3! 3⋅2⋅1

There are 20 possible captain/co-captain pairs.

5. How many different ways can the letters in the word “SUCCESS” be
rearranged?

Solution:

23
Since the order matters, we have to calculate the permutations. There are
7 letters we can choose from, and 7 spots to put them in.

n! 7! 7! 7 ⋅ 6 ⋅ 5 ⋅ 4 ⋅ 3 ⋅ 2 ⋅ 1
n Pk = = = = = 5,040
(n − k)! (7 − 7)! 0! 1

But since the letter S repeats three times in the word and the letter C
repeats twice, the actual number of unique rearrangements will be less
than 5,040. We have to divide by 3! for the S and by 2! for the C.

5,040 5,040 5,040

= = = 420
3! ⋅ 2! 3 ⋅ 2 ⋅ 1 ⋅ 2 ⋅ 1 12

There are 420 possible arrangements of the letters.

6. Mrs. B’s kindergarten class has 14 students and Mr. G’s kindergarten
class has 16 students. Three students will be selected at random from each
of these classrooms to ride on a float in the school parade coming up next
week. How many different groups of 6 can be chosen to ride the float?

Solution:

Since order doesn’t matter, this is a combination question. We need to find

the combination for Mrs. B’s class, and then the combination for Mr. G’s
class. Then we’ll multiply those together get the total number of
combinations.

nCk ⋅ nCk = 14C3 ⋅ 16C3 = 364 ⋅ 560 = 203,840

24
BINOMIAL RANDOM VARIABLES

1. We toss a fair coin 15 times and record the number of tails. Is this
experiment modeled by a binomial random variable? If it isn’t, explain why.
If it is, determine its parameters n and p and express the binomial random
variable as X ∼ B(n, p).

Solution:

Yes, this experiment results in a binomial random variable. Let X be the

number of tails observed out of 15 tosses. We know that p = 0.5 for each
trial because there are only two possible outcomes, heads or tails.
Therefore, X ∼ B(15,0.5).

2. We randomly select students from our school until we find a student

in the school band. Assume there are 900 students in the school and 80
participate in the school band. Is this experiment modeled by a binomial
random variable? If it isn’t, explain why. If it is, determine its parameters n
and p and express the binomial random variable as X ∼ B(n, p).

Solution:

No, this experiment does not result in a binomial random variable. We do

have a fixed probability of success,

25
 80 4
p= = ≈ 0.09
900 45

and the trials can be considered independent because we have a large

population. But we’re not using a fixed number of trials, because we’re
continuing to select students until we find one in the band, and we don’t
know how many trials that will take.

3. Let X ∼ B(n, p) be a binomial random variable with n = 12 and p = 0.08.

Find P(X = 4).

Solution:

We’re being asked to find the probability that we get exactly 4 successes
in 12 trials, if the probability of success is p = 0.08.

(4)
12
P(X = 4) = (0.08)4(1 − 0.08)8

P(X = 4) = (495)(0.08)4(1 − 0.08)8

P(X = 4) ≈ 0.0104

4. Let Y be the number of times we roll a 1 on a fair 6-sided die if we do

10 trials. Fill in the following probability distribution for Y, rounding each
probability to 4 decimal places.

26
 Y 0 1 2 3 4 5 6 7 8 9 10

P(Y)

Solution:

With n = 10, p = 1/6, and k = 0, 1, 2, 3, . . . 10, find P(Y = k) for each value of k
using

(k)
n
P(k successes in n trials) = p k(1 − p)n−k

After rounding each value to 4 decimal places, the table is

Y 0 1 2 3 4 5 6 7 8 9 10

P(Y) 0.16 0.32 0.29 0.16 0.05 0.01 0 0 0 0 0

5. For each binomial random variable, determine whether the shape of

the probability distribution will be skewed right, skewed left, or
symmetrical.

1. X ∼ B(n, p) with n = 10 and p = 0.15

2. Y ∼ B(n, p) with n = 10 and p = 0.75

3. Z ∼ B(n, p) with n = 10 and p = 0.50

27
Solution:

The probability distribution for X will be skewed right because the

probability of success, p = 0.15, is less than 0.5.

The probability distribution for Y will be skewed left because the

probability of success, p = 0.75, is greater than 0.5.

The probability distribution for Z will be symmetrical because the

probability of success, p = 0.50, is exactly 0.5.

6. Suppose an environmental biologist is studying juvenile sunfish

mortality. He finds that only 30 % of juvenile sunfish survive in a certain
lake. Out of 8 randomly selected juvenile sunfish, what is the probability
that exactly 3 will survive?

Solution:

We’re finding the probability that we get exactly 3 successes in 8 trials.

(3)
8
P(X = 3) = (0.3)3(1 − 0.3)5

P(X = 3) = (56)(0.3)3(1 − 0.3)5

P(X = 3) ≈ 0.2541

28
POISSON DISTRIBUTIONS

1. A student is able to solve 10 practice problems per hour, on average.

Find the probability that she can solve 12 in the next hour.

Solution:

We know this is a Poisson experiment with the following values:

λ = 10, the average number of practice problems solved in an hour

x = 12, the number of homework problems she wants to complete in

the next hour

The Poisson probability is

λ xe −λ
P(x) =
x!

1012e −10
P(12) =
12!

P(12) ≈ 0.0948

So the probability the student will solve 12 homework problems is

approximately 0.0948 or 9.48 % .

29
 2. A student is able to solve 6 practice problems per hour, on average.
Find the probability that she can solve at least 4 in the next hour.

Solution:

The probability that the student solves at least 4 practice problems is the
probability that she doesn’t solve either 0, 1, 2, or 3 practice problems. So
the probability we need to find is

P(X ≥ 4) = 1 − P(X = 0) − P(X = 1) − P(X = 2) − P(X = 3)

We know this is a Poisson experiment with λ = 6, the average number of

practice problems solved in an hour, so the Poisson probability is

60e −6 61e −6 62e −6 63e −6

P(X ≥ 4) = 1 − − − −
0! 1! 2! 3!

e −6 6e −6 36e −6 216e −6
P(X ≥ 4) = 1 − − − −
1 1 2 6

1 6 18 36
P(X ≥ 4) = 1 − − − − 6
e 6 e 6 e 6 e

1
P(X ≥ 4) = 1 − (1 + 6 + 18 + 36)
e 6

61
P(X ≥ 4) = 1 −
e6

P(X ≥ 4) ≈ 0.8488

30
So the probability the student will solve at least 4 homework problems is
approximately 0.8488 or 84.88 % .

3. A student is able to solve 5 practice problems per hour, on average.

Find the probability that she solves at most 3 in the next hour.

Solution:

The probability that the student solves at most 3 practice problems is the
probability that she solves either 0, 1, 2, or 3 practice problems. So the
probability we need to find is

P(X ≤ 3) = P(X = 0) + P(X = 1) + P(X = 2) + P(X = 3)

We know this is a Poisson experiment with λ = 5, the average number of

practice problems solved in an hour, so the Poisson probability is

P(X ≤ 3) = P(X = 0) + P(X = 1) + P(X = 2) + P(X = 3)

50e −5 51e −5 52e −5 53e −5

P(X ≤ 3) = + + +
0! 1! 2! 3!

e −5 5e −5 25e −5 125e −5
P(X ≤ 3) = + + +
1 1 2 6

1 5 25 125
P(X ≤ 3) = 5 + 5 + 5 + 5
e e 2e 6e

( 6 )
1 25 125
P(X ≤ 3) = 1 + 5 + +
e 5 2

31
 e (6 6 )
1 36 75 125
P(X ≤ 3) = 5 + +
6

118
P(X ≤ 3) = 5
3e

P(X ≤ 3) ≈ 0.2650

So the probability the student will solve at most 3 homework problems is

approximately 0.2650 or 26.50 % .

4. A baker is able to bake 50 loaves of bread per day, on average. Find

the probability that he can bake 60 on Friday.

Solution:

We know this is a Poisson experiment with the following values:

λ = 50, the average number of loaves of bread baked per day

x = 60, the number of loaves of bread he wants to bake on Friday

The Poisson probability is

λ xe −λ
P(x) =
x!

5060e −50
P(60) =
60!

32
 P(60) ≈ 0.0201

So the probability the baker will bake 60 loaves of bread is approximately

0.0201 or 2.01 % .

5. A baker is able to bake 10 cakes per hour, on average. Find the

probability that he can bake more than 5 in the next hour.

Solution:

The probability that the baker bakes more than 5 cakes is the probability
that he doesn’t bake either 0, 1, 2, 3, 4, or 5 cakes. So the probability we
need to find is

P(X > 5) = 1 − P(X = 0) − P(X = 1) − P(X = 2)

−P(X = 3) − P(X = 4) − P(X = 5)

We know this is a Poisson experiment with λ = 10, the average number of

cakes baked in an hour, so the Poisson probability is

100e −10 101e −10 102e −10

P(X > 5) = 1 − − −
0! 1! 2!

103e −10 104e −10 105e −10

− − −
3! 4! 5!

e −10 10e −10 100e −10

P(X > 5) = 1 − − −
1 1 2

33
 1,000e −10 10,000e −10 100,000e −10
− − −
6 24 120

1 10 50 500 1,250 2,500

P(X > 5) = 1 − − − − − −
e 10 e 10 e 10 3e 10 3e 10 3e 10

( 3 )
1 500 1,250 2,500
P(X > 5) = 1 − 1 + 10 + 50 + + +
e 10 3 3

4,433
P(X > 5) = 1 −
3e 10

P(X > 5) ≈ 0.9329

So the probability the baker will bake more than 5 cakes is approximately
0.9329 or 93.29 % .

6. A baker is able to frost 2 cakes per hour, on average. Find the

probability that he frosts fewer than 5 cakes in the next hour.

Solution:

The probability that the baker frosts fewer than 5 cakes is the probability
that he frosts either 0, 1, 2, 3, or 4 cakes. So the probability we need to find
is

P(X < 5) = P(X = 0) + P(X = 1) + P(X = 2) + P(X = 3) + P(X = 4)

We know this is a Poisson experiment with λ = 2, the average number of

cakes frosted in an hour, so the Poisson probability is

34
 P(X < 5) = P(X = 0) + P(X = 1) + P(X = 2) + P(X = 3) + P(X = 4)

20e −2 21e −2 22e −2 23e −2 24e −2

P(X < 5) = + + + +
0! 1! 2! 3! 4!

e −2 2e −2 4e −2 8e −2 16e −2
P(X < 5) = + + + +
1 1 2 6 24

1 2 2 4 2
P(X < 5) = 2 + 2 + 2 + 2 + 2
e e e 3e 3e

e2 ( 3 3)
1 4 2
P(X < 5) = 1 + 2 + 2 + +

7
P(X < 5) =
e2

P(X < 5) ≈ 0.9473

So the probability the baker will frost fewer than 5 cakes is approximately
0.9473 or 94.73 % .

35
“AT LEAST” AND “AT MOST,” AND MEAN, VARIANCE, AND STANDARD
DEVIATION

1. Assume X is a binomial random variable. Let X ∼ B(n, p) with n = 15 and

p = 0.45. Find P(X > 7).

Solution:

Since we’re running n = 15 trials, and we want to find the probability that
we get the first success after the 7th trial, we can express this as

P(X > 7) = P(X = 8) + P(X = 9) + . . . + P(X = 15)

which is the same as

1 − P(X ≤ 7)

Find P(X ≤ 7).

(0) (1)
15 0 15 15
P(X ≤ 7) = (0.45) (1 − 0.45) + (0.45)1(1 − 0.45)14

(2) (3)
15 15
+ (0.45)2(1 − 0.45)13 + (0.45)3(1 − 0.45)12

(4) (5)
15 15
+ (0.45)4(1 − 0.45)11 + (0.45)5(1 − 0.45)10

(6) (7)
15 15
+ (0.45)6(1 − 0.45)9 + (0.45)7(1 − 0.45)8

36
 P(X ≤ 7) = 0.5515 + 15(0.45)(0.5514)

+105(0.452)(0.5513) + 455(0.453)(0.5512)

+1,365(0.454)(0.5511) + 3,003(0.455)(0.5510)

+5,005(0.456)(0.559) + 6,435(0.457)(0.558)

P(X ≤ 7) ≈ 0.6535

Then 1 − P(X ≤ 7) is

1 − P(X ≤ 7) ≈ 1 − 0.6535

1 − P(X ≤ 7) ≈ 0.3465

2. According to a 2017-2018 survey, 68 % of U.S. households own a pet.

Suppose we select 12 households at random. What is the probability that
fewer than 8 of them own a pet?

Solution:

Let X be the number of households that own a pet. Then we can express
the variable as X ∼ B(12,0.68). The probability that we’ll have fewer than 8
successes is

P(X < 8) = P(X ≤ 7) = P(X = 0) + P(X = 1) + . . . + P(X = 7)

Find P(X < 8).

37
 (0) (1)
12 0 12 12
P(X < 8) = (0.68) (1 − 0.68) + (0.68)1(1 − 0.68)11

(2) (3)
12 12
+ (0.68)2(1 − 0.68)10 + (0.68)3(1 − 0.68)9

(4) (5)
12 12
+ (0.68)4(1 − 0.68)8 + (0.68)5(1 − 0.68)7

(6) (7)
12 6 6 12
+ (0.68) (1 − 0.68) + (0.68)7(1 − 0.68)5

P(X < 8) = (0.3212) + 12(0.68)(0.3211)

+66(0.682)(0.3210) + 220(0.683)(0.329)

+495(0.684)(0.328) + 792(0.685)(0.327)

+924(0.686)(0.326) + 792(0.687)(0.325)

P(X < 8) ≈ 0.3308

3. According to a 2017-2018 survey, 68 % of U.S. households own a pet.

Suppose 200 households are selected at random. Find the expected value
and standard deviation for the number of households that own a pet.

Solution:

38
Let X be the number of households that own a pet. Then we can express
the variable as X ∼ B(12,0.68). The expected value is

μX = E(X ) = (200)(0.68) = 136 households

The variance is

σX2 = Var(X ) = (200)(0.68)(1 − 0.68) = 43.52

which means the standard deviation is

σX = SD(X ) = 43.52 ≈ 6.597 households

4. 3 % of runners in the Boston Marathon do not finish. Suppose we

select a SRS of 140 Boston Marathon runners. How many do we expect to
finish the race?

Solution:

Let X be the number of runners who finish the Boston Marathon. Then we
can say X ∼ B(140,0.97). Then the expected value is

μX = E(X ) = (140)(0.97) = 135.8 runners

5. We roll a fair die 6 times. What is the probability we’ll observe an even
number in at most 3 of the rolls?

39
Solution:

Let X be the number of even numbers observed. Then we can say

X ∼ B(6,0.5).

P(X ≤ 3) = P(X = 0) + P(X = 1) + P(X = 2) + P(X = 3)

(0) (1)
6 6
P(X ≤ 3) = (0.5)0(1 − 0.5)6 + (0.5)1(1 − 0.5)5

(2) (3)
6 2 4 6
+ (0.5) (1 − 0.5) + (0.5)3(1 − 0.5)3

P(X ≤ 3) = (0.56) + 6(0.5)(0.55) + 15(0.52)(0.54) + 20(0.53)(0.53)

P(X ≤ 3) = 0.56 + 6(0.56) + 15(0.56) + 20(0.56)

P(X ≤ 3) = 42(0.56)

P(X ≤ 3) ≈ 0.6563

6. We roll two fair 6-sided die 10 times and observe the sum. What is the
probability of rolling a sum of 7 on at least six of the rolls?

Solution:

Let X be the number of times we roll a sum of 7. Since there are 36 possible
rolls when we roll two die, and 6 of them result in a sum of 7,

40
 1 2 3 4 5 6

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

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

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

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

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

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

the probability is

6 1
P(sum of 7) = =
36 6

Therefore we can express X as

( 6)
1
X ∼ B 10,

So the probability of rolling a sum of 7 at least 6 times out of 10 rolls is

P(X ≥ 6) = P(X = 6) + P(X = 7) + P(X = 8) + P(X = 9) + P(X = 10)

or

1 − P(X ≤ 5)

Find P(X ≤ 5).

( 0 )(6) ( 6) ( 1 )(6) ( 6)
0 10 1 9
10 1 1 10 1 1
P(X ≤ 5) = 1− + 1−

41
 ( 2 )(6) ( 6) ( 3 )(6) ( 6)
2 8 3 7
10 1 1 10 1 1
+ 1− + 1−

( 4 )(6) ( 6) ( 5 )(6) ( 6)
4 6 5 5
10 1 1 10 1 1
+ 1− + 1−

(6) (6)(6)
10 9
5 1 5
P(X ≤ 5) = + 10

(6) (6) (6) (6)

2 8 3 7
1 5 1 5
+45 + 120

(6) (6) (6) (6)

4 6 5 5
1 5 1 5
+210 + 252

P(X ≤ 5) ≈ 0.9976

Then 1 − P(X ≤ 5) is

1 − P(X ≤ 5) ≈ 1 − 0.9976

1 − P(X ≤ 5) ≈ 0.0024

or about a 0.24 % chance.

42
BERNOULLI RANDOM VARIABLES

1. A game at the local county fair involves spinning a circular spinner

that’s divided into 8 congruent sections, only two of which are “winners.”
We buy 5 spins for $3.00. If we land on “winner” on any of our 5 spins, we
get to choose a stuffed animal. Is this an example of Bernoulli trials?

Solution:

The set of 5 spins can be considered Bernoulli trials because the spins are
independent of one another, there are exactly two outcomes (land on a
winning, or not), and the probability of success (landing on a winner)
remains constant for each trial at p = 2/8 = 1/4 = 0.25 = 25 % .

2. A game at the local county fair involves spinning a circular spinner

that’s divided into 8 congruent sections, only two of which are “winners.”
We buy 5 spins for $3.00. If we land on “winner” on any of our 5 spins, we
get to choose a stuffed animal. Find the mean and standard deviation for
each trial.

Solution:

43
We already know that the probability of winning on any single spin is
p = 2/8 = 1/4 = 0.25 = 25 % , which means μ = p = 0.25. The standard
deviation will therefore be

σ= p(1 − p) = (0.25)(1 − 0.25) = 0.1875 ≈ 0.4330

3. A game at the local county fair involves spinning a circular spinner

that’s divided into 8 congruent sections, only two of which are “winners.”
We buy 5 spins for $3.00. If we land on “winner” on any of our 5 spins, we
get to choose a stuffed animal. Find the mean and standard deviation for
the number of winners expected in a set of 5 spins.

Solution:

We already know that the probability of winning on any single spin is

p = 2/8 = 1/4 = 0.25 = 25 % , which means μ = p = 0.25. Therefore, for 5 spins
the mean will be μ = np = 5(0.25) = 1.25. And the standard deviation for 5
spins will be

σ= np(1 − p) = (5)(0.25)(1 − 0.25) = 0.9375 ≈ 0.9682

4. A game at the local county fair involves spinning a circular spinner

that’s divided into 8 congruent sections, only two of which are “winners.”
We buy 5 spins for $3.00. If we land on “winner” on any of our 5 spins, we

44
get to choose a stuffed animal. Find the probability of observing no
winners in a set of 5 spins.

Solution:

The probability of spinning a winner is

2 1
P(winner) = p = = = 0.25
8 4

P(non-winner) = 1 − p = 1 − 0.25 = 0.75

Therefore, the probability of no winners in 5 spins is

P(no winners in 5 spins) = (0.75)5 = 0.2373

5. A game at the local county fair involves spinning a circular spinner

that’s divided into 8 congruent sections, only two of which are “winners.”
We buy 5 spins for $3.00. If we land on “winner” on any of our 5 spins, we
get to choose a stuffed animal. What is the probability of observing at
least 1 winner in a set of 5 spins?

Solution:

If we observe at least one winner out of 5 spins, that means we’re looking
for the probability of getting 1, 2, 3, 4, or 5 winners. The only result we’re
excluding is the probability of getting 0 winners. Which means we could

45
flip this problem around and calculate the probability of at least 1 winner
as

P(W ≥ 1) = 1 − P(W = 0)

P(W ≥ 1) = 1 − 0.2373

P(W ≥ 1) = 0.7627

6. Our goal is to learn about the percentage of students with high ACT
scores. We randomly select high school seniors and record their highest
ACT score. Explain why these aren’t Bernoulli trials. Then design a way to
conduct the experiment differently so that they can be considered
Bernoulli trials.

Solution:

These are not Bernoulli trials because actual ACT scores are recorded. This
is a random variable, but the variable can take on many different values,
not simply “success” or “failure.”

To change the experiment so that we’re running Bernoulli trials, we could

define a specific range of ACT scores as “failures” and another range as
“successes.” For instance, we could define a success as a score of 28 or
higher, and a failure as a score lower than 28 (27 or lower).

46
Then the probability of a senior having a score of 28 or higher will have
some constant probability of success from trial to trial, so we now have an
experiment in which we’re using Bernoulli trials.

47
GEOMETRIC RANDOM VARIABLES

1. We toss a coin until we get “tails.” Does this experiment represent a

geometric random variable? If it doesn’t, explain why. If it does, determine
its parameter p and express the variable as X ∼ Geom(p).

Solution:

Yes, this experiment results in a geometric random variable. Let X be the

number of trials it takes to get our first “tails.” We know that p = 0.5 for
each trial because there are two equally likely outcomes when we flip a
coin. So X ∼ Geom(0.5).

2. We randomly select students from our school until we find a student

in the school band. Assume there are 900 students in the school and 80
participate in the school band. Does this experiment represent a
geometric random variable? If it doesn’t, explain why. If it does, determine
its parameter p and express the variable as X ∼ Geom(p).

Solution:

Yes, this experiment results in a geometric random variable. We do have a

fixed probability of success,

48
 80 4
p= = ≈ 0.09
900 45

and the trials can be considered independent because we have a large

population. We’re selecting students until we find someone in the band.
Therefore X ∼ Geom(0.09).

3. Let X ∼ Geom(p) with p = 0.25. Find P(X = 5).

Solution:

We’re being asked to find the probability that we get our first success on
the 5th trial, if the probability of success on any single trial is p = 0.25.

P(X = n) = p(1 − p)n−1

P(X = 5) = (0.25)(1 − 0.25)5−1

P(X = 5) = (0.25)(0.75)4

P(X = 5) ≈ 0.0791

4. Suppose we roll a 6-sided fair die until we observe a 2. What is the

probability that a 2 will be observed within the first 5 trials?

Solution:

49
The probability of success on any single trial is p = 1/6, which means the
probability of failure is 1 − p = 1 − (1/6) = 5/6. Therefore, the probability that
we get a 2 within the first 5 trials is

P(X ≤ 5) = P(X = 1) + P(X = 2) + P(X = 3) + P(X = 4) + P(X = 5)

6 (6) 6 (6) 6 (6) 6 (6) 6 (6)

1−1 2−1 3−1 4−1 5−1
1 5 1 5 1 5 1 5 1 5
P(X ≤ 5) = + + + +

6 (6) 6 (6) 6 (6) 6 (6) 6 (6)

0 1 2 3 4
1 5 1 5 1 5 1 5 1 5
P(X ≤ 5) = + + + +

P(X ≤ 5) ≈ 0.5981

5. Suppose we roll a 6-sided fair die until we observe a 2. What is the

probability that a 2 won’t be observed until at least the 6th trial?

Solution:

The probability of success on any single trial is p = 1/6, which means the
probability of failure is 1 − p = 1 − (1/6) = 5/6. Therefore, the probability that
we get a 2 within the first 5 trials is

P(X ≤ 5) = P(X = 1) + P(X = 2) + P(X = 3) + P(X = 4) + P(X = 5)

6 (6) 6 (6) 6 (6) 6 (6) 6 (6)

1−1 2−1 3−1 4−1 5−1
1 5 1 5 1 5 1 5 1 5
P(X ≤ 5) = + + + +

50
 6 (6) 6 (6) 6 (6) 6 (6) 6 (6)
0 1 2 3 4
1 5 1 5 1 5 1 5 1 5
P(X ≤ 5) = + + + +

P(X ≤ 5) ≈ 0.5981

Therefore, the probability that we don’t observe a success until the 6th
trial or later is

P(X ≥ 6) ≈ 1 − 0.5981

P(X ≥ 6) ≈ 0.4019

6. According to a 2017-2018 survey, 68 % of U.S. households own a pet.

Suppose we start randomly surveying households and asking whether
they are pet owners. How many do we expect we will need to survey to
find our first household that owns a pet?

Solution:

Let X be the trial when we find our first pet owner. We know that X is a
geometric random variable with X ∼ Geom(0.68). Then the expected value
is

1 1
μX = E(X ) = = ≈ 1.471
p 0.68

So we could say that we expect we’ll need to survey somewhere between

1 and 2 households in order to find our first pet-owning household.

51
52