Chapter 3:

Probability Distributions
Stokes S. Baker, MS, PhD
Definitions
Variable: Any item being measured
Random variable: Any variable whose value are controlled by an
element of chance.
Discrete random variable: Random variable with discrete values
Continuous random variable: Random variable with continuous
random variable.
Capital Lower
letter case
letter

Review of notation: P(# = %)

Random Value of the
variable random variable
 Sec. 3.1: Probability distribution
Probability distribution: (a.k.a., density function)
An accounting of the probability of all possible outcomes of an discrete
random variable.
Figure 3.1, Roll of 2 dice

●
● 0.18 NOTE:
0.16 %
1/6 x 1/6 = 0.0278 0.14
P(X = x) 0.12 ! P ' = ) = 1.0
0.1 "#$
0.08
0.06
0.04
Probabilities are
0.02 exhausted
0
1 2 3 4 5 6 7 8 9 10 11 12 13
Sum of roll
Sec 3.1: Binomial Distributions
Binomial events: Only two possible outcomes
Gender: Boy versus Girl
Disease state: Healthy vs. Diseased

Binomial distribution: Theoretical probability distribution of binomial outcomes.

Assumptions:
1. Fixed number of trials (observations)
2. Independent events
3. Probability is constant
Room of three people
According to US Census, only 23.6% of Michiganders have a college degree
• Event A: College degree Hyperlink
• P(X) = 0.236
Binomial probability equation
Where:
n = number of trials
k = No. of time the event occurs and n > k
p = probability of the event occurring

n!
p (1 - p )
n -k
P( X = k ) = k

k! ( n - k )!
ænö n!
n Ck = ç
ç k ÷÷ = k! (n - k )!
è ø
 Numerical example:
What is the probability that a room with 3 Michiganders, 2 will have college
degrees. Remember, 23.6% of Michiganders have a college degree.
n = number of trials = 3 people
k = No. of time the event occurs = 2 with degrees
p = probability of the event occurring = 0.236

P( X = k ) =
n!
p k (1 - p )
n -k P ( X = 2) = 3(0.0557)(0.764 )
k! ( n - k )!
3! P ( X = 2) = 0.1277
(0.263) 2 (1 - 0.263)
3- 2
P ( X = 2) =
2! (3 - 2)!

3 ´ 2!
(0.263) (0.764 )
1
P ( X = 2) = 2

2!´1!
Excel’s binomial probability function
=BINOMDIST(k,n,p,FALSE)
Expected value, mean and standard deviation of a
binomial distribution
Expected value: The most likely outcome = μ

Population mean: E(x) = μ = np

where n = number of trials and p is the probability.

Population standard deviation: ! = #$ % − $

The shape of the binomial distribution is determine
by n and p Link

Fig. 3.3 Fig. 3.4 Fig. 3.5

Concept Check 3.1
In urban setting, 30% of grey squirrels have black fur.
a) If 18 squirrels are captured at random, how many squirrels do you expect
to have black fur?
b) What is the probability that exactly 4 of the 18 squirrels will have black fur?
c) What is the probability that at least 1 of the 18 squirrels will have black fur?

a) n = 18 squirrels, p = 0.30; E(x) = μ = np = 18 x 0.30 = 5.4 squirrels

b) n = 18 squirrels, k = 4 black, p = 0.30;
n!
p k (1 - p )
n -k
P( X = k ) =
k! ( n - k )!
18!
P ( X = 4) = (0.30) (1 - 0.30)
4 18- 4

4! (18 - 4)!
18 ´ 17 ´ 16 ´ 15 ´ 14!
P ( X = 4) = (0.0081)(0.00678) = 0.1681
(4 ´ 3 ´ 2 ´ 1) ´ (14!)
Concept Check 3.1
In urban setting, 30% of grey squirrels have black fur.
a) If 18 squirrels are captured at random, how many squirrels do you expect
to have black fur?
b) What is the probability that exactly 4 of the 18 squirrels will have black fur?
c) What is the probability that at least 1 of the 18 squirrels will have black fur?
c) n = 18 squirrels, p = 0.30, k > 0 18!
P ( X = 0) = (1) ´ (0.70)18
P(X > 0) = P(X = 0)C 1 ´ (18! )
n!
p (1 - p )
n -k
P( X = k ) = P ( X = 0) = (0.70 ) = 0.001628
k 18

k! ( n - k )!
P ( X = 0) =
18!
(0.30)0 (1 - 0.30)18-0 P ( X = 0)C = (1 - 0.001628) = 0.998372
0! (18 - 0)!
Sec. 3.3: Poisson Distribution
Poisson distribution: Distribution of rare evens.
• Developed by Simeon Denis Poisson, an 18th Century French
mathematician
• An approximation of the binomial distribution
Assumptions:
• Binomial probability (i.e., constant P, fixed number of trials)
• P is small
• n is large
Sec. 3.3: Poisson Distribution
n!
Binomial probability equation: p (1 - p )
n -k
P( X = k ) = k

k! ( n - k )!

% &' ()
Poisson equation: P "=$ =
$!
k = number of times the event occurs
e = 2.71828
λ = μ = n·p
n = number of trials
p = binomial probability
 Poisson distribution can be a good approximation of
the binomial distribution
When:
• p is small
• n is large
 Poisson distribution can be a poor approximation of
the binomial distribution
When:
• p is large
• n is small
Numerical example:
Typical cell bacterial colony forms after 30 doubling from a
single cell.
• How many cells in a colony? (2)30 = 1.07 x 109 cells

Spontaneous formation of rifampicin resistant cells

(RifR) is 1.5 x 10-8 Creative commons copyright, Beatrice
Leung and Shijun Liu

Q1: How many RifR cells do you expect to find in a typical

bacterial colony?

E(x) = μ = np = (1.07 x 109) (1.5 x 10-8) = 1.605 x 101 ≈ 16 colonies

Numerical example:
Typical cell bacterial colony contains (2)30 = 1.07 x 109 cells
Rifampicin resistant (RifR) = 1.5 x 10-8 (RifR) per cell division
E(x) = μ = np = (1.07 x 109) (1.5 x 10-8) = 16.05 cells
Q2: What is the probability of finding exactly 16 RifR cells
in a colony?
Creative commons copyright, Beatrice
Leung and Shijun Liu

&' ) k = 16
% (
P "=$ = ; e = 2.71828
$!
P(X = k) = [e λ ] / k! λ =-λ k
μ = 16.05
(2.71828)/01.34 (56.78)01
P " = 16 = = 0.099209797
56!
 Numerical example:
Typical cell bacterial colony contains (2)30 = 1.07 x 109 cells
Rifampicin resistant (RifR) = 1.5 x 10-8 (RifR) per cell division
E(x) = μ = np = (1.07 x 109) (1.5 x 10-8) = 16.05 cells
Q3: What is the probability of at least 1 RifR cells in a colony?

Creative commons copyright, Beatrice

P(X > 0) = P(X = 0)C Leung and Shijun Liu

k=0
% &' () ; e = 2.71828
P "=$ =
P(X = k) = [e λ ] / k!
$! λ -λ
= μ =k16.05
(2.71828)./0.23 (45.67)2
P "=0 = = 1.07 x 10-7
6!
P(X > 0) = P(X = 0)C = 1 – 1.07 x 10-7 = 0.999
Excel binomial and Poisson functions
=[Link](k,n,p,FALSE)
λ=n·p
=[Link](k,λ,FALSE)
Sec. 3.4: Continuous random variable
• Continuous variable controlled by random events
• Y-axis on frequency distribution graphs: f(x)
Sec. 3.5: Binomial distribution can approximate a
normal distribution
When P = 0.5 and n is relatively large Link

• Unimodal
• Symmetrical
When n = ∞ the binomial distribution becomes a
normal curve
 Normal curve describe by μ and σ
2
1 æ x-µ ö
1 - ç
2è s ø
÷
f ( x) = e
2ps

Any normal curve describe by:

•μ
•σ
 Empirical rule
Figure 3.11

1σ
F(x)

σ
68.3 %
95.5 %
99.7 %
μ and σ defines the normal curve
μ=0
Figure 3.12 μ=1 Figure 3.13
μ=2
σ=1
σ=2
σ=3

When μ changes: When σ changes:

• Location changes • Shape changes
• Shape does not change • Location does not change
Sec. 3.6: Standard normal distribution
Standard normal distribution is a normal distribution where:
• μ=0
• σ=1

f(x)

σ’s from μ
 Converting any normal curve into a
standard normal curve
#−% Where: Z = No. of σ from μ
!= μ = population mean

Link
& σ = population standard deviation
x = data value
μ=5 xi/σ
σ=2 xi - µ

Excel file: Standard normal slider

Area of right tail of the standard normal curve
Standard normal distribution is a normal distribution, μ = 0 and σ = 1
Z = No. of σ from μ #−%
μ = population mean !=
σ = population standard deviation &
x = data value

100.0 %
-68.3
68.3 %
31.7% f(x)
31.7% / 2 = 15.86 %

15
μ

. 86

.8
6%
15
Z
Right tail of the standard normal curve
Table A3, Page A-9

x-µ
Z =
s

Right tail
What is the area of the right tail 1.96 σ from μ?

F(x)

?
μ
Z 1.96 Z
Area of the right tail 1.96 σ from μ?
#−%
!=
&
Area of the right tail 1.96 σ from μ
#−%
!=
&

Answer: 0.02500
Using the standard normal distribution
In the USA, the mean height of women is 161 cm with a standard deviation of 9 cm. The
population is normally distributed. What proportion of the population is above 175 cm?
1) Draw a sketch of the question.
2) Determine the value of the
known variables
3) Determine value of the unknown
variable
Z = No. of σ from μ
μ = population mean = 161 cm
σ = population standard deviation = 9 cm
x = data value =175 cm
#$% '() *+ $',' *+
!= = = '. ),
& - *+
 Using the standard normal distribution
In the USA, the mean height of women is 161 cm with a standard deviation of 9 cm. The
population is normally distributed. What proportion of the population is above 175 cm?

Z = No. of σ from μ
μ = population mean = 161 cm
σ = population standard deviation = 9 cm
x = data value =175 cm
#$% '() *+ $',' *+
! = = = 1.56
& - *+

Answer: 0.05938 or 5.938%

What proportion of the ♀ population > 150 cm?
μ = population mean = 161 cm #$% '()*+ $','*+
σ = population standard deviation = 9 cm ! = & = -*+
= −1.22
x = data value =150 cm Since –Z, in left tail: 0.11123 or 11.123%
STEPS:
1) Draw a sketch of the question.
2) Determine the value of the known variables
3) Determine value of the unknown variable
What proportion of the ♀ population < 150 cm?
μ = population mean = 161 cm
σ = population standard deviation = 9 cm
x = data value =150 cm
STEPS:
1) Draw a sketch of the question.
2) Determine the value of the known variables
3) Determine value of the unknown variable

23
.11
? =0 P(X < 150) = P(X > 150)C = 1 – P(X < 150)
= 1 – 0.1123 = 0.8877
 How tall is the top 2.5% of the ♀ population?
μ = population mean = 161 cm
σ = population standard deviation = 9 cm
STEPS:
1) Draw a sketch of the question.
2) Determine the value of the known variables
3) Determine value of the unknown variable

0.025
Z = 1.96 x = σZ + µ
#$%
!= x = 9cm ∙1.96 + 161cm
&
x – µ = σZ x = 178.64 cm
What values describe the middle 95% of the population?
μ = population mean = 161 cm
σ = population standard deviation = 9 cm
The population is normally distributed.

x1 = 9cm x -1.96Z + 161 cm

x1 = 143.36 cm
Z1 = -1.96; Z2 = 1.96 x2 = 9cm x 1.96Z + 161 cm
#$% x2 = 178.64 cm
!= x = σZ + µ
&