Chapter 2: Special Probability Distributions

Kavikumar PhD

October 1, 2018
Outline

Introduction

Binomial Distribution

Poisson Distribution
Poisson Approximation to the Binomial Distribution

The Normal Distribution

The Standard Normal Distribution
Normal Approximation to Binomial
Concept Mapping

Probability
Distribution

Discrete Random Continuous Random

Variable Variable

Binomial Hypergeometric Poisson Normal

Distribution Distribution Distribution Distribution

Standard Normal
Distribution
Binomial Distribution
Binomial Distribution is a basic distribution that models chance
variation in repetitions of an experiment that has only two possible
outcomes.
Experiment

Success Failure

p q

Probability : p+q =1
Binomial Experiment: The experiment consists of repeating a trial
over certain number of times. Each trial had two possible
outcomes.
Binomial Distribution B(n, p)

A coin is tossed n times. Assume P(H) = p, so that

P(T ) = q = (1 − p). Assume the trials are independent. Let X
denote the number of H’s in n trials. Then X ∈ {0, 1, 2, · · · , n}.
Its probability distribution is
  
 n
p x q n−x , x = 0, 1, 2, · · · , n
P(X = x) = x

0, otherwise.
 
n
Note that binomial coefficient is defined as
x
 
n n!
=
x r !(n − r )!
Binomial Distribution B(n, p)

A binomial experiment assumes the following conditions:

I The experiment consists of n trials (n is fixed in advance).
I There are only two possible outcomes on each trial: S (for
success) and F (for failure).
I The trials are identical in the sense that P(H) = p for each
trial.
I The trials are independent, so the outcomes of any particular
trial does not affect the other outcomes.
I The binomial random variable X is the number of successes in
n trials and written as X ∼ B(n, p).
ULLI TRIALS AND BINOMIAL DISTRIBUTIONS
Histogram
The histogram of the binomial distribution is:
am of the binomial distribution is given in Figure 18.1
 Histogram: Example
Construct the binomial distribution for the total number of heads
ial distribution
in four flips of a is givencoin.
balanced byMake
theafollowing
histogram. table

r
Solution: The binomial 0distribution
1 is2given3by the4 following table
1 4 6 4 1
p(r) x 16 016 1 162 16 3 416
1 4 6 4 1
P(X = x) 16 16 16 16 16
ponding histogram is shown in Figure 19.1
The corresponding histogram is shown in Figure:
Problem 1
Suppose that in a box of 100 computer chips, the probability of a
chip to be defective is 3%. Inspection process for defective chips
consists of selecting with replacement 5 randomly chosen chips in
the box and to send the box for shipment if none of the five chips
is defective. Write down the random variable, the corresponding
probability distribution and then determine the probability that the
box described here will be allowed to be shipped.

Solution: Let X be the number of defective chips in the box.

Then X is a binomial random variable with probability distribution
 
5
P(X = x) = (0.03)x (0.97)5−x , x = 0, 1, 2, 3, 4, 5.
x

Now,

P(sheet goes into circulation) = P(X = 0) = (0.97)5 = 0.859.

Problem 2
Suppose that 40% of the voters in a city are in favor of a ban of
smoking in public buildings. Suppose 5 voters are to be randomly
sampled. Find the probability that
1. 2 favor the ban.
2. less than 4 favor the ban.
3. at least 1 favor the ban.
Solution:  
5
1. P(X = 2) = (0.4)2 (0.6)3 ≈ 0.3456.
2
2.
P(X < 4) = P(X = 0) + P(X = 1) + P(X = 2) + P(X = 3)
   
5 0 5 5
= (0.4) (0.6) + (0.4)1 (0.6)4
0 1
   
5 2 3 5
+ (0.4) (0.6) + (0.4)3 (0.6)2 ≈ 0.913
2 3
 
5
3. P(X ≥ 1) = 1 − P(X < 1) = 1 − (0.4)0 (0.6)5 ≈ 0.922.
0
Problem 3
A student takes a test consisting of 10 true-false questions.
1. What is the probability that the student answers at least six
questions correctly?
2. What is the probability that the student answers at most two
questions correctly?
Solution:
1. Let X be the number of correct responses. Then X is a
binomial random variable with parameters n = 10 and
p = 0.5. So, the desired probability is
P(X ≥6) = P(X = 6) + P(X = 7) + · · · + P(X = 10) =
P10 10
x=6 (0.5)x (0.5)10−x ≈ 0.3769.
x
 
P2 10
2. P(X ≤ 2) = x=0 (0.5)x (0.5)10−x ≈ 0.0547.
x
Problem 4

A study shows that 30 percent people aged 50-60 in a certain town

have high blood pressure. What is the probability that in a sample
of fourteen individuals aged between 50 and 60 tested for high
blood pressure, more than six will have high blood pressure?

Solution: Let X be the number of people in the town aged 50-60

with high blood pressure. Then X is a binomial random variable
with n = 14 and p = 0.3. Thus,

P(X > 6) = 1 − P(X ≤ 6)

X6  
14
=1− (0.3)i (0.7)14−i
i
i=0
≈ 0.0933.
Mean and Variance of Binomial Distribution

If X ∼ B(n, p), then X = X1 + X2 + · · · + Xn where Xi ’s are

independent variables with P(Xi = 1) = p, 1 ≤ i ≤ n.
i.e., the binomial distribution with n trials and probability of
success, p has

Mean, µ = E (X ) = E (X1 + X2 + · · · + Xn ) = np,

Variance, σ 2 = Var (X ) = Var (X1 + X2 + · · · + Xn ) = npq,
√
Standard deviation, (SD), σ = npq.
Problem 1

The probability of a student passing an exam is 0.2. Ten students

took the exam.
1. What is the probability that at least two students passed the
exam?
2. What is the expected number of students who passed the
exam?
3. How many students must take the exam to make the
probability at least 0.99 that a student will pass the exam?
Solution:. Let X be the number of students who passed the exam.
Then, X has a binomial distribution with n = 10 and p = 0.2.
Problem 1
1. The event that at least two students passed the exam is
{X ≥ 2}. So,
P(X ≥ 2) = 1 − P(X < 2) = 1 − P(X = 0) − P(X = 1)
   
10 0 10 10
=1− (0.2) (0.8) − (0.2)1 (0.8)9
0 1
≈ 0.6242.
2. E (X ) = np = (10).(0.2) = 2.
3. Suppose that n students are needed to make the probability at
least 0.99 that a student will pass the exam. Let A denote the
event that a student pass the exam. Then, Ac means that all
the students fail the exam. We have,
P = 1 − P(Ac ) = 1 − (0.8)n ≥ 0.99.
ln(0.01)
Solving the inequality, we find that n ≥ ≈ 20.6. So,
ln(0.8)
the required number of students is 21.
Problem 2

Let X be a binomial random variable with parameters (12, 0.5).

Find the variance and the standard deviation of X .

Solution:.
We have n = 12 and p = 0.5.
Thus, Var (X ) = np(1 − p) = 6(1
√ − 0.5) = 3.
The standard deviation is σ = 3.
Example 1
Suppose a tennis player makes a successful first serve 70% of the
time. Assume that each service is independent of the others.
Assume he serves 80 times in a match:
1. What are the mean and SD of the number of good first serves
expected?
2. Verify that you can use a normal mode to fit the distribution
of the number of good serves.

Proof.
The first serve is a Bernoulli trial, with P(Good Serve)=0.7. Let
X =number of good serves in 80 times. Then X ∼ B(80, 0.7).
Then
1. µ = np = 80(0.70) √
= 56 and
√
SD(X ) = npq = 56 × 0.3 = 4.10
2. Since, np = 56 ≥ 10, nq = 24 ≥ 10, B(80, 0.7) may be
approximated by N(56, 4.10).
Example 2

The probability that a patient recovers from a delicate heart

operation is 0.9. What is the probability that exactly five out of
the seven patients having this operation will survive?
Proof.
Let X be the number of patient that recovers from heart operation
of 7 patients. When n = 7 and p = 0.9 we have
 
7
P(X = x) = (0.9)x (0.1)7−x , x = 0, 1, 2, · · · , 7
x

P(X = 5) = P(X ≤ 5) − P(X ≤ 4) = 0.1497 − 0.0257 = 0.1240

Example 3
In a certain city district the need for money to buy drugs is stated
as the: reason for 75% of all thefts. Find the probability that
among the next 5 theft cases reported in this district.
1. exactly 2 resulted from the need for money for buy drugs.
2. at most 3 resulted form the need for money to buy.
Proof.
Let X be the number of thefts resulted from the need for money to
buy drugs of 5 theft cases. When n = 5 and p = 3/4 we have
   x  5−x
5 3 1
P(X = x) = , x = 0, 1, 2, · · · , 5
x 4 4
   2  3
5 3 1
1. P(X = 2) = = 0.0879
2 4 4
P
2. P(X ≤ 3) = 3x=0 P(X = x) = 1 − P(X = 4) − P(X = 5)
   4  1    5  0
5 3 1 5 3 1
=1− − = 0.3672.
4 4 4 5 4 4
Example 4
In testing a certain kind of truck tire over a rugged terrain, it is
found that 25% of the trucks fail to complete the test run without
a blowout. Of the next 15 trucks tested, find the probability that
1. from 3 to 6 have blowouts,
2. fewer than 4 have blowouts,
3. more than 5 have blowouts.
Proof.
Let X be the number of times that blowout of 15 times. For
n = 15, and p= 0.25,
 we have
15
P(X = x) = (0.25)x (0.75)15−x , x = 0, 1, 2, · · · , 15
x
1. P(3 ≤ X ≤ 6) = P(X ≤ 6) − P(X ≤ 2) = 0.9434 − 0.2361 =
0.7073.
2. P(X < 4) = P(X ≤ 3) = 0.4613.
3. P(X > 5) = 1 − P(X ≤ 5) = 1 − 0.8516 = 0.1484.
Example 5
It is known that 60% of mice inoculated with a serum are
protected from a certain disease. If 5 mice are inoculated, find the
probability that
1. none contracts the disease,
2. fewer than 2 contract the disease,
3. more than 3 contract the disease.

Proof.
Let X be the number of mice that contract the disease of 5 mice
inoculated. For
 n=  5, and p = 0.4, we have
5
P(X = x) = (0.4)x (0.6)5−x , x = 0, 1, 2, · · · , 5
x
1. P(X = 0) = 0.0778.
2. P(X < 2) = P(X ≤ 1) = 0.3370.
3. P(X > 3) = 1 − P(X ≤ 3) = 1 − 0.9130 = 0.0.870.
Example 6
In a study of brand recognition, 95% of consumers recognized
Coke. The company randomly selects 4 consumers for a taste test.
Let X be the number of consumers who recognize Coke of 4
consumers.
1. Write out the PMF table for this.
2. Find the probability that among the 4 consumers, 2 or more
will recognize Coke.
3. Find the expected number of consumers who will recognize
Coke.
4. Find the variance for the number of consumers who will
recognize Coke.
Solution: For n = 4 and p = 0.95, we have
 
4
P(X = x) = (0.95)x (0.05)4−x , x = 0, 1, 2, 3, 4
x
Solution
1.
X   p(x)
4
0 (0.95)0 (0.05)4 = 0.0000062
0
 
4
1 (0.95)1 (0.05)3 = 0.00047
 1
4
2 (0.95)2 (0.05)2 = 0.01354
 2
4
3 (0.95)3 (0.05)1 = 0.17147
 3
4
4 (0.95)4 (0.05)0 = 0.81451
4
2.
P(X ≥ 2) = P(X = 2) + P(X = 3) + P(X = 4)
= 0.01354 + 0.17147 + 0.81451
= 0.99952
Solution

3.

E (X ) = 0(0.0000062) + 1(0.00047) + 2(0.01354)

+ 3(0.17147) + 4(0.81451) = 3.8

or
E (X ) = np = 4 × 0.95 = 3.8
4.
Var (X ) = np(1 − p) = 4 × (0.95) × (0.05) = 0.19
Example 7
A company installs new central heating furnaces, and has found
that for 15% of all installations a return visit is needed to make
some modifications. Six installations were made in a particular
week. Assume independence of outcomes for these installations.
1. What is the probability that a return visit was needed in all of
these cases?
2. What is the probability that a return visit was needed in none
of these cases?
3. What is the probability that a return visit was needed in more
than one of these cases?
Solutions:
1.
 
6
P(X = 6) = (0.15)6 (0.85)0
6
= 1 × (0.0000114) × 1 = 0.0000114
Example 7

2.
 
6
P(X = 0) = (0.15)0 (0.85)6
0
= 1 × (1) × 0.3771 = 0.3771

3.

P(X > 1) = 1 − [P(X = 0) + P(X = 1)]

   
6 1 5
= 1 − 0.3771 + (0.15) (0.85)
1
= 1 − [0.3771 + 6 × 0.15 × 0.4437]
= 1 − (0.3771 + 0.3993)
= 0.2236
Example 8

A company that produces fine crystal from experience that 10% of

its goblets have cosmetic flaws and must be classified as
”seconds”.
1. Among six randomly selected goblets, how likely is it that only
one is a second?
2. Among six randomly selected goblets, what is the probability
that at least two are seconds?
3. If goblets are examined one by one, what is the probability
that at most five must be selected to find four that are not
seconds?
Example 8
Solution: Note X ∼ Bin(6, 0.10). Then
 
6
1. P(X = 1) = (0.1)1 (0.9)5 = 0.3543
1
2. P(X ≥ 2) = 1 − [P(X = 0) + P(X = 1)].
From (1), P(X
 =1) = 0.3543 and
6
P(X = 0) = (0.1)0 (0.9)6 = 0.5314.
0

P(X ≥ 2) = 0.3543 + 0.5314 = 0.1143.

3. Either 4 or 5 goblets must be selected:
3.1 select 4 goblets
 with zero defects:
4 0 4
P(X = 0) = (0.1) (0.9) = 0.6561
0
3.2 select 4 goblets, one of which has a defect and the 5th is good:
 
4 1 3
(0.1) (0.9) × 0.9 = 0.26244
1
So, the required probability is 0.6561+0.26244=0.91854.
Example 9
Twenty percent of all telephones of a certain type are submitted
for service while under warranty. Of these, 60% can be repaired,
whereas the other 40% must be replaced with new units. If a
company purchases ten of these telephones, what is the probability
that exactly two will end up being replaced under warranty?
Solution:
Let S represent a telephone that is submitted for service while
under warranty and must be replaced. Then

p = P(S) = P(replaced|submitted)P(submitted) = (0.40)(0.20) = 0.08.

Thus X , the number among the company’s 10 phones that must

be replaced, has a binomial distribution with n = 10, p = 0.08.
Hence,
 
10
p(2) = P(X = 2) = (0.08)2 (0.92)8 = 0.1478.
2
Example 10
A newsboy purchases papers at 10 cents and sells them at 15 cents.
However, he is not allowed to return unsold papers. Suppose his
daily demand is a binomial random variable with n = 10, p = 1/3.
1. How many papers should he purchase to maximize his
expected profit?
2. Comment briefly on the use of the binomial distribution as a
probability model in this situation.
Solution: The daily demand, D is B(10, 1/3). Let Y denote the
number of papers sold during the day, and b denote the number of
papers he purchased.
Consider possible b’s.
y 0 1 2
I When b = 2, ⇒ profit -20 -5 10
P(Y = y ) 0.017 0.087 0.896
where, for y < b,
   y  10−y
10 1 2
P(Y = y ) = P(X = y ) = and
y 3 3
E [profit] = −(−20)(0.017) + (−5)(0.087) + (10)(0.896) = 8.185.
Example 10
y 0 1 2 3
I When b = 3, ⇒ profit -30 -15 0 15
P(Y = y ) 0.017 0.087 0.195 0.701
and

E [profit] = (−30)(0.017)+(−15)(0.087) + (0)(0.195)

+ (15)(0.701) = 8.700

I When b = 4,
y 0 1 2 3 4
profit -40 -25 -10 5 20 and
P(Y = y ) 0.017 0.087 0.195 0.260 0.441

E [profit] = (−40)(0.017)+(−25)(0.087) + (−10)(0.195)

+ (5)(0.260) + (20)(0.441) = 5.315

We have found a local maximum of expected profit at b = 3. It is

fairly clear this is also the global maximum.
Example 11
Let X be a binomial
 random variablen+1with parameters n and p.
1 1 − (1 − p)
Show that E = .
X +1 (n + 1)p
Solution: Let X denote a binomial random variable with
parameters n and p. then
  X n  
1 1 n
E = p k (1 − p)n−k
X +1 k +1 k
k=0
n
X n!
= p k (1 − p)n−k
(n − k)!(k + 1)!
k=0
Xn
1 (n + 1)!
= p k+1 (1 − p)n−k
(n + 1)p (n − k)!(k + 1)!
k=0
" n #
1 X (n + 1)!
= p j (1 − p)n−j+1 − (1 − p)n+1
(n + 1)p (n − j + 1)!j!
k=0
1
= [1 − (1 − p)n+1 ]
(n + 1)p
Example 12

Let X be a binomial random variable with parameters n and p.

What value of p maximizes P(X = k) for a given value of k in
{0, 1, · · · , n}?
Solution:
 
n
P(X = k) = p k (1 − p)n−k
k
 
d n
P(X = k) = k p k−1 (1 − p)n−k
dp k
 
n
− (n − k) p k (1 − p)n−k−1 = 0
k
k
p=
n
Poisson Distribution
I A Poisson distribution is a tool that helps to predict the
probability of certain events from happening when you know
how often the event has occurred.
I It gives us the probability of a given number of events
happening in a fixed interval of time. For example:
1. Number of accidents at a particular place.
2. Number of tornadoes in a year.
3. Number of people affected by a disease.
Such phenomena, with small probability of occurrence (called rare
events), can be described by Poisson distribution. Here the random
variable X denotes the number of such events during a certain
time period. Note X ∈ {0, 1, · · · }. The pmf of X is

e −λ λx
P(X = x) = , x = 0, 1, · · ·
x!
where: λ is mean number of event.
x is number of success event.
Mean and Variance of the Poisson Distribution
The mean E (X ) of the Poisson distribution is
∞
X ∞
X e −λ λx−1 X e λ λx ∞
e −λ λx
E (X ) = x. =λ =λ = λe −λ e λ = λ.
x! (x − 1)! x!
x=0 x=1 x=0

The variance Var (X ) of the Poisson distribution is

∞
X X∞
e −λ λk e −λ λk−2
E (X (X − 1)) = k(k − 1) = λ2 k=2
k! (k − 2)!
k=2
∞
X e −∞ λk
=λ 2
= λ2 e −λ e λ = λ2
k!
k=0

Then, we find

E (X 2 ) = E (X (X − 1)) + E (X ) = λ2 + λ.

Thus,
Var (X ) = E (X 2 ) − (E (X ))2 = λ.
Problem 1

The number of car accidents on a certain section of highway I40

averages 2.1 per day. Assuming that the number of accidents in a
given day follows a Poisson distribution, what is the probability
that 4 accidents will occur on a given day?

Solution:
The probability that 4 accidents will occur on a given day is given
by
(2.1)4
P(X = 4) = e −2.1 ≈ 0.0992.
4!
Problem 2
The number of people entering a movie theatre averages one every
two minutes. Assuming that a Poisson distribution is appropriate.
1. What is the probability that no people enter between 12:00
and 12:05?
2. Find the probability that at least 4 people enter during
[12:00,12:05].
Solution:
1. Let X be the number of people that enter between 12:00 and
12:05. We model X as a Poisson random variable with
parameter λ, the average number of people that arrive in the
5-minute interval. But, if 1 person arrives every 2 minutes, on
average (so 1/2 a person per minute), then in 5 minutes an
average of 2.5 people will arrive. Thus, λ = 2.5. Now

2.50
P(X = 0) = e −2.5 = e −2.5 .
0!
Problem 2

2.

P(X ≥ 4) = 1 − (PX ≤ 3)
= 1 − P(X = 0) − P(X = 1) − P(X = 2) − P(X = 3)
2.50 2.51 2.52 2.53
= 1 − e −2.5 − e −2.5 − e −2.5 − e −2.5
0! 1! 2! 3!
Problem 3
The number of weekly life insurance sold by an insurance agent
averages 3 per week. Assuming that this number follows a Poisson
distribution, calculate the probability that in a given week the
agent will sell
1. some policies
2. 2 or more policies but less than 5 policies.
3. Assuming that there are 5 working days per week, what is the
probability that in a given day the agent will sell one policy?
Solution:
1. Let X be the number of policies sold in a week. Then
e −3 30
P(X ≥ 1) = 1 − P(X = 0) = 1 − ≈ 0.95021.
0!
2.
P(2 ≤ X < 5) = P(X = 2) + P(X = 3) + P(X = 4)
e −3 32 e −3 33 e −3 34
= + + ≈ 0.61611.
2! 3! 4!
Problem 3

3. Let X be the number of policies sold per day. Then

λ = 3/5 = 0.6. Thus,

e −0.6 (0.6)
P(X = 1) = ≈ 0.32929.
1!
Problem 4

The number X of people in a certain town who get the influenza

follows a Poisson distribution. The proportion of people who did
not get the flu is 0.01. Find the probability mass function of X .
e −λ
Solution: The pmf of X is P(X = k) = λk , k = 0, 1, 2, · · · .
k!
−λ
Since P(X = 0) = 0.01 we can write e = 0.01. Thus, λ = 4.605
and the exact Poisson distribution is
e −4.605
P(X = k) = (4.605)k , k = 0, 1, 2, · · ·
k!
Problem 5
Misprints in a book averages one misprint per 10 pages. Suppose
that the number of misprints is a random variable having Poisson
distribution. Let X denote the number of misprints in a stack of
50 pages. Find the mean and the standard deviation of X .

Solution: Since the book has 1 misprint per 10 pages, the number
of misprints in a stack of 50 pages is 5. Thus, X is a Poisson
random variable with parameter

λ = 5.

Hence,
E (X ) = λ = 5
and √
σ= 5.
Problem 6
Let X represent the number of customers arriving during the
morning hours and let Y represent the number of customers
arriving during the afternoon hours at a diner. You are given:
1. X and Y are Poisson distributed.
2. The first moment of X is less than the first moment of Y by
8.
3. The second moment of X is 60% of the second moment of Y .
Calculate the variance of Y .
Solution: We have
E (X ) = E (Y ) − 8 ⇒ E (X )2 = E (Y )2 − 16E (Y ) + 64. But
E (X 2 ) = Var (X ) + E (X )2 = E (X ) + E (X )2 = 0.6E (Y 2 ) =
0.6(E (Y ) + E (Y )2 ). Hence
0.6(E (Y ) + E (Y )2 ) = E (Y )2 − 16E (Y ) + 64 yields the quadratic
equation 0.4E (Y )2 − 15.5E (Y ) + 56 = 0 whose roots are
E (Y ) = 4 and E (Y ) = 35. The value E (Y ) = 4 yields
E (X ) = 4 − 8 = −4 which is impossible since the expected value of
a Poisson random variable is always positive. Hence, E (Y ) = 35.
Problem 7
Suppose hurricanes arrive with a mean of 2.35 per year. Assuming
Poisson model,
1. What is the probability that no hurricanes next year?
2. What is the probability that during the next 2 years, there is
exactly one hurricane?
Solution: Let X =number of hurricanes in a year. Given,
X ∼ P(2.35). Hence, E (X ) = λ = 2.35.
e −2.35 (2.35)0
1. P(no hurricane next year) = P(X = 0) = =
0!
0.095.
2.
P(exactly one hurricane in next 2 year)
= P(one in I year) × P(no hurricane in II year)
+ P(no hurricane in I year) × P(one hurricane in 2 year)
= P(X1 = 1)P(X2 = 0) + P(X1 = 0)P(X2 = 1)
= 2.e −4.70 (2.35) = 0.04347,
Problem 8

In a study of a drug-induced anaphylaxis among patients taking

rocuronium bromide as part of their anaesthesia, Laake and
Rottingen found that the occurrence of anaphylaxis followed a
Poisson model with λ = 12 incidents per year in Norway. Find
1. The probability that in the next year, among patients receiving
rocuronium, exactly three will experience anaphylaxis?
2. The probability that less than two patients receiving
rocuronium, in the next year will experience anaphylaxis?
3. The probability that more than two patients receiving
rocuronium, in the next year will experience anaphylaxis?
4. The expected value of patients receiving rocuronium, in the
next 6 months who will experience anaphylaxis.
5. The variance and standard deviation of patients receiving
rocuronium, in the year who will experience anaphylaxis.
Solution Problem 8
Let X be a number o patients that have anaphylaxis from
medication per year and λ is an average of patients that have
anaphylaxis from medication per year=12 per year. Then
e −λ λx
P(X = x) = , x = 0, 1, 2, · · ·
x!
e −12 123
1. P(X = 3) = = 0.001769533
3!
2. P(X <  2)0= P(X1  ≤ 1) = P(X = 0) + P(X = 1) =
12 12
e −12 + = 0.0000799
0! 1!
3. P(X > 2) = 1 − P(X ≤ 2) = 1 − [P(X =0) + P(X =
12 2
1) + P(X = 2)] = 1 − 0.0000799 − e −12 = 0.999478
2!
4. µ = 6 = E (X ) = λ
5. σ 2 = 12 = Var (X ) = λ
√ p
6. σ = 12 = Var (X )
Problem 9

Let X have a Poisson distribution, with probability mass function

e −λ λk
p(k) = , and let Y = 2−X . Find the expected value of Y .
k!
Solution:
X λk −λ
E (Y ) = 2−k e
k!
k
X (λ/2)k
= e −λ
k!
k
−λ λ/2
=e e
= e −λ/2
Poisson Approximation to the Binomial Distribution

If random variable X is binomially distributed with its parameter n

and p that fulfil the condition the n ≥ 30 and p < 0.1, then the
binomial distribution can be approximated to Poisson distribution
with µ = np, i.e.,

X ∼ B(n, p) → X ∼ P(x, µ)
Problem 1
In a group of 100 individuals, let X be the random variable
representing the total number of people in the group with a
birthday on Thanksgiving day. Then X is a binomial random
variable with parameters n = 100 and p = 1/365. What is the
probability at least one person in the group has a birthday on
Thanksgiving day?

Solution: We have
    
100 1 0 364 100
P(X ≥ 1) = 1−P(X = 0) = 1− ≈ 0.2399.
0 365 365

Using the Poisson approximation, with

1 100 20
λ = 100 × = = we find
365 365 73
(20/73)0 − 20
P(X ≥ 1) = 1 − P(X = 0) = 1 − e 73 ≈ 0.2396.
0!
Problem 2
Consider the experiment of rolling two fair dice 6 times. Let X
denote the number of times a double 4 appears. For k = 0, 1, 2
compare P(X = k) found using the binomial distribution with the
one found using Poisson approximation.
Solution: We are given that n = 6 and p = 1/36 so that
λ = np = 1/6. Let B(X = k) be the probability using binomial
distribution and P(X = k) be the probability using Poisson
distribution. Then
 6
1
B(X = 0) = 1− ≈ 0.8445 ⇔ P(X = 0) = e −1/6 ≈ 0.8465
36
   5  
6 1 1 −1/6 1
B(X = 1) = 1− ≈ 0.1448 ⇔ P(X = 1) = e
1 36 36 6
≈ 0.1411
  2  4
6 1 1
B(X = 2) = 1− ≈ 0.0103 ⇔ P(X = 1) =
2 36 36


1 2
−1/6 6
e ≈ 0.0118
2!
Problem 3
Consider a contest where a participant fires at a small can placed
on the top of box. Each time the can is hit, it is replaced by
another can. Suppose that the probability of a participant hitting
the can is 1/32 . Assume that the participant shoots 96 times, and
that all shoots are independent.
1. Find the probability mass function of the number of shoots
that hit a can.
2. Give an approximation for the probability of the participant
hitting no more than one can.
Solution:
1. Let X denote the number of shoots that hit a can. Then X is
binomially distributed:
n
P(X = k) = ( )p k (1 − p)n−k , n = 96, p = 1/32.
k
2. Since n is large, and p small, we can use the Poisson
approximation, with parameter λ = np = 3. Thus, P(X ≤
1) = P(X = 0) + P(X = 1) ≈ e −λ (1 + λ) = 4e −3 ≈ 0.199.
Problem 4
Theorem
If X is a Poisson random variable with parameter λ, then
λ
P(X = k + 1) = P(X = k).
k +1

Proof.
We have
λk+1
e −λ
P(X = k + 1) (k + 1)!
=
P(X = k) λk
e −λ
k!
λ
=
k +1
Normal Distribution

I Random variables can be either discrete or continuous.

I A discrete random variable cannot assume all values between
any two given values of the variables.
I A continuous random variable can assume all values between
any two given values of the variable.

Definition
A random variable X has a normal distribution and it is referred to
as a normal random variable if and only if its probability density is
given by
1 1 x−µ 2
N(x; σ 2 ) = √ e − 2 ( σ ) for − ∞ < x < ∞
σ 2π
where σ > 0.
ntinuous Random Variables and Distributions
Normal Density Curves
rmal Density Curves

µ = 0; =1
µ = 0; = 0.6
0.6 µ = 0; =2
µ = 2; = 0.8

0.4

0.2

0
6 4 2 0 2 4 6
Example

Let X be a normal random variable with mean 950 and standard

deviation 10. Find P(947 ≤ X ≤ 950).

Solution: We have
Z 950
1 (x−950)2
P(947 ≤ X ≤ 950) = √ e− 200 dx ≈ 0.118.
10 2π 947
The Standard Normal Distribution
Definition
Standard Normal Distribution is that normal variable, X transform
into Standard Normal Variable, Z with mean, µ = 0 and standard
deviation σ = 1.
X −µ
If X ∼ N(µ, σ 2 ), then Z = has a standard normal
σ
N(0, 1) distribution.

X ∼ N(µ, σ 2 ) −→ Z ∼ N(0, 1).

Standartization
The Standard score or Z score
X −µ
Z=
σ

N(µ, σ 2 ) −−−−−−−−−−−−−−−−−−→ N(0, 1)

The cumulative distribution function of Z as

Φ(Z ) = P(Z ≤ z).

     
a−µ b−µ b−µ a−µ
P(a ≤ X ≤ b) = P ≤Z ≤ =Φ −Φ
σ σ σ σ
 Continuous Random Variables and Distributions
Percentiles of the Standard Normal Distribution
Percentiles of the Standard Normal Distribution

z curve

Shaded area = 0.99

Continuous Random Variables and Distributions

0 z0.01

99th percentile
z curve

Shaded area = 0.05

(P. Vellaisamy: MSU-STT-351-Sum-17B)
Shaded area = 0.05
Probability & Statistics for Engineers 39 / 49

1.645 = z0.05 0 1.645 = z0.05

5th percentile 95th percentile

P(Z > a) = P(Z ≥ a)=from table

Let zα be such that P(Z > zα ) = α. That is, zα is the

[100(1 − α)]th percentile of N(0, 1). For example z1 = 0.1587.
P(Z < −a) = P(Z > a)=from table

For example P(Z < −1) = 0.1587 = P(X > 1)

=
P(Z < a) = P(Z ≤ a) = 1 − P(Z < a)
For example P(Z < 1) = 1 − 0.1587 = 0.8413

−
P(Z > −a) = 1 − P(Z < a)
For example P(Z > −1) = 1 − 0.1587 = 0.8413

−
P(|Z | > a) = 2 × P(Z > a)

For example P(|Z | > a) = 0.3174 and 2 × P(Z > a) = 2 × 0.1587

= 2×
P(|Z | < a) = 1 − 2 × P(Z > a)
For example P(|Z | < 1) = 1 − 2 × 0.1587 = 0.6827

−2×
P(a < Z < b) = P(Z > a) − P(Z > b)
P(0.25 < Z < 1.5) = 0.3345,
P(Z > 0.25) − P(Z > 1.5) = 0.4013 − 0.0668 = 0.3345

−
P(−a < Z < −b) = P(Z > b) − P(Z > a)
P(−1.5 < Z < −0.25) = 0.3345,
P(Z > 0.25) − P(Z > 1.5) = 0.4013 − 0.0668 = 0.3345

−
P(−a < Z < b) = 1 − P(Z > a) − P(Z > b)
P(−1 < Z < 2) = 0.8186,
1 − P(Z > 1) − P(Z > 2) = 1 − 0.1587 − 0.0228 = 0.8186

− −
Problem 1

The reaction time for an in-traffic response to a brake signal from

standard brake lights can be modeled as X ∼ N(1.25, 0.462 ). Then
 
1 − 1.25 1.75 − 1.25
P(1 ≤ X ≤ 1.75) = P ≤Z ≤
0.46 0.46
= P(−0.54 ≤ Z ≤ 1.09)
= Φ(1.09) − Φ(−0.54)
= 0.5675
Problem 2
If X is a normal random variable with mean 80 and standard
deviation 10, compute the following probabilities by standardizing:
1. P(X ≤ 100)
2. P(65 ≤ X ≤ 100)
3. P(|X − 80| ≤ 10).
Solution:  
100 − 80
1. P(X ≤ 100) = P z ≤ = P(Z ≤ 2) = Φ(2.00) =
10
0.9772  
65 − 80 100 − 80
2. P(65 ≤ X ≤ 100) = P ≤Z ≤ =
10 10
P(−1.5 ≤ Z ≤ 2) = Φ(2.00) − Φ(−1.50) =
0.9772 − 0.0668 = 0.9104.
3. P(|X − 80| ≤ 10) = P(−10 ≤ X − 80 ≤ 10) = P(70 ≤ X ≤
90) = P(−1 ≤ Z ≤ 1) = Φ(1) − Φ(−1) = 0.6826.
ontinuous Random Variables and Distributions
Normal
ormal Approximation
approximation to Binomial
to Binomial

0.20 B (20, 0.4)

N(8, 2.19)

0.15

0.10

0.05

0.00
0 5 10 15 20
x
Normal Approximation to Binomial

Let X ∼ B(n, p).

If np ≥ 10 and n(1 − p) ≥ 10, then X has approximately a normal
distribution with
µ = np
and
√
σ= npq
and
 
area under the normal curve
P(X ≤ x) = B(n, p) ≈
to the left of x + 0.5
 
x + 0.5 − np
=Φ √ , x = 0, 1, 2 · · · , n
npq
Problem 1

Suppose 25% of all licensed drivers do not have insurance. Let X

be the number of uninsured drivers in a random sample of size 50.
Then X ∼ B(50, 0.25) and

P(5 ≤ X ≤ 15) = P(X ≤ 15) − P(X ≤ 4)

   
15.5 − 12.5 4.5 − 12.5
≈Φ −Φ
3.06 3.06
= 0.8320.
Problem 2
Suppose that 10% of all steel shafts produced by a certain process
are nonconforming but can be reworked (rather than having to be
scrapped). Consider a random sample of 200 shafts, and let X
denote the number among these that are nonconforming and can
be reworked. What is the (approximate) probability that X is
between 15 and 25 (inclusive)?

Solution:

P(15 ≤ X ≤ 25) = P(X ≤ 25) − P(X ≤ 14)

   
25 + 0.5 − 20 14 + 0.5 − 20
=Φ √ −Φ √
18 18
= Φ(1.3) − Φ(−1.3)
= 0.9032 − 0.0968 = 0.8064.
Problem 3
Suppose only 75% of all drivers in a state regularly wear a seat
belt. A random sample of 500 drivers is selected. Find the
probability that
1. Between 360 and 400 (inclusive) of the drivers in the sample
regularly wear a seat belt?
2. Fewer than 400 of those in the sample regularly wear a seat
belt?
Solution:
1. Given n = 500, p = 0.75 ⇒ µ = 375, σ = 9.68246. Using
normal approximation,

P(360 ≤ X ≤ 400) = P(359.5 ≤ X ≤ 400.5)

= P(−1.60 ≤ Z ≤ 2.58)
= 0.9409

2. P(X < 400) = P(X < 399.5) = P(Z ≤ 2.53) = 0.9943.

Problem 4

In a box of 100 light bulbs, 10 are found to be defective. What is

the probability that the number of defectives exceeds 13?

Solution: Let X be the number of defective items. Then X is

binomial with n = 100 and p = 0.1. Since np = 10 > 5 and
nq = 90 > 5 we can use the normal approximation to the binomial
with µ = np = 10 and σ 2 = npq = 9. We want P(X > 13). Using
continuity correction we find

P(X > 13) = P(X ≥ 14)

 
X − 10 13.5 − 10
=P √ ≥ √
9 9
≈ 1 − Φ(1.17) = 1 − 0.8790 = 0.121.
Problem 5
In a small town, it was found that out of every 6 people 1 is
left-handed. Con- sider a random sample of 612 persons from the
town, estimate the probability that the number of left handed
persons is strictly between 90 and 150.

Solution: Let X be the number of left-handed people in the

sample. Then X is a binomial random variable with n = 612 and
p = 1/6. Since np = 102 > 5 and n(1 − p) = 510 > 5 we can use
the normal approximation to the binomial with µ = np = 102 and
σ 2 = np(1 − p) = 85. Using continuity correction we find

P(90 < X < 150) = P(91 ≤ X ≤ 149)

 
90.5 − 102 X − 102 149.5 − 102
=P √ ≤ √ ≤ √
85 85 85
= P(−1.25 ≤ Z ≤ 5.15)
≈ 0.8943.
Problem 6
There are 90 students in a statistics class. Suppose each student
has a standard deck of 52 cards of his/her own, and each of them
selects 13 cards at random without replacement from his/her own
deck independent of the others. What is the chance that there are
more than 50 students who got at least 2 aces ?

Solution: Let X be the number of students who got at least 2

aces or more, then clearly X is a binomial random variable with
n = 90 and
4 C2 · 48 C11 4 C3 · 48 C10 4 C4 · 48 C9
p= + + ≈ 0.2573
52 C13 52 C13 52 C13

Since np ≈ 23.157 > 5 and n(1 − p) ≈ 66.843 > 5, X can be

approximated
p by a normal random variable with µ = 23.157 and
σ = np(1 − p) ≈ 4.1473.Thus,
 
50.5 − 23.157
P(X > 50) = 1−P(X ≤ 50) = 1−Φ ≈ 1−Φ(6.59).
4.1473