Chapter 5

Discrete Probability
Distributions

David Chow
Sep 2021

1
Learning Objectives

In this chapter, you will learn:

 The properties of a probability distribution (PD)
 To compute the expected value and variance of a PD

 To solve problems by the binomial distribution

 The sections below are NOT covered in this course

 Covariance and its application in finance
 Poisson distribution
 Hypergeometric distribution

2
Random Variables

 A random variable (rv) represents

a possible numerical value from
an uncertain event

 A random variable can be discrete or continuous

 Eg: No. of boys in a 2-children family
 Eg: Weight of a randomly-selected CU student

3
Random Variables
Definition: A random variable (rv) represents a
possible numerical value from an uncertain event

Examples
- Value of a dice
- Waiting time for MTR

4
 Probability Distribution
 A probability distribution for a discrete random variable
is a mutually exclusive listing of all possible numerical
outcomes and the probability associated with each outcome
 Eg: X = # heads when a fair coin is tossed twice

5
Eg: Probability Distribution
 Two conditions for a discrete PD:
(i) f(x) ≥ 0, and
(ii) Σf(x) = 1

x = No. of cars sold during

a day at Motor One

6
 Expected Value
 A probability distribution has ____ properties to
describe: ________
 The mean () of a probability distribution is the
expected value of its random variable

  = E(X) N X P(X )

i i
 where N = no. of all possible outcomes
 X1, X2, …, XN denote values of possible outcomes

 E(X) is an average of all possible outcomes weighted

by the respective probability

7
Eg: No. of Boys
 Find E(X) where X = no. of boys in a 2-children family
 First, the probability distribution

X = no. of boys P(X)

0 0.25
1 0.50
2 0.25
 E(X) =
 Interpretation
 E(X) =1 here is the average outcome when the process is
repeated many times

8
 Eg: Credit Card
 The table below shows the number of credit cards owned by a US adult:
 Number of Credit Cards (X) P(X)
 0 0.26
 1 0.22
 2 0.16
 3 0.12
 4 0.08
 5 0.06
 6 0.04
 7 0.03
 8 0.02
 9 0.01
 Compute and interpret E(X)

ANS: E(X) = 2.22

_____
9
 Interpretation of E(X): Rolling a Dice

 X = possible outcomes = 1, 2, …, 6
 P(X) = ⅙ for all Xs
 E(X) = ∑ X∙P(X) = 3.5 -- Why is it the average value of X?

 Let’s look at the formula this way:

 E(X) = (1)(⅙ ) + … + 6(⅙ )

 Think of the probability (⅙ ) as the result of rolling

a dice many times, say, 6,000 times
 Then we have a data set of 6,000 values
 How many of them are “1”s?

 E(X) = (1,000x1 + 1,000x2 +…+1,000x6) / 6,000

10
Measuring Dispersion
 Variance of a discrete random variable
N
σ   [X i  E(X)]2 P(X i )
2

i 1

 Standard deviation of a discrete random variable

N
σ σ  2
 [X
i 1
i  E(X)]2 P(X i )

where: E(X) = Expected value of the discrete random variable X

Xi = the ith outcome of X
P(Xi) = Probability of the ith occurrence of X

11
 Measuring Dispersion

 Eg2: In the credit card example, σ = 2.1890

______

12
 Application of Expected Values:
Lottery Games

 Mega Millions is a popular US lottery game

 There was a jackpot of $521 million in 2018

 It’s the fourth-biggest prize ever

 Ticket price is $2

Q: Is it worth buying?

13
 Application of Expected Values:
Lottery Games

 Let’s compute E(X)

 Define X as the profit from a ticket – prize net of ticket prize

 We need to know the prizes, and the probabilities of

winning different prizes, then we can find E(X)
 It turns out that E(X) = -$0.03
 The average profit from a ticket is negative!

Q: Is it worth buying?

14
Binomial Distribution

15
 The Binomial Distribution

 Define X = no. of successes in n trials

 Eg: X = no. of heads if a coin is tossed 15 times
 Eg: X = no. of defective bulbs per 10 new bulbs

 X is a binomial random variable

IF it has certain properties
 Then P(X) can be found by the binomial formula

16
Properties of the Binomial Dist
 Properties
 A fixed number (n) of identical observations
 Eg: 15 tosses of a coin (n=15),
 Eg: 10 light bulbs from a production line (n=10)

Identical observations?
You may think of it as
repeated draws (with replacement)

17
 Properties of the Binomial Dist
Properties (cont)
 Two mutually exclusive & collectively exhaustive
categories, “success” and “failure”
 If P (Success) = π, then P (Fail) = ____

Examples
 People who have “taken” or “not yet taken” vaccinations
 New job applicants either accept or reject the offer

Our interest is in the no. of successes in n trials

18
 Example: MC Quiz
 You are randomly guessing on a quiz of 5 MCs
 Each MC has 5 choices
 What is the probability of getting a zero?

 Answer
 Define Success = correct answer, then P(S) = 0.2 i.e., π=0.2
 5 identical & independent trials i.e.,
n=5
 X = no. of success = 0, 1, 2, 3, 4, or 5
 P (zero score) = P(X=0) = (0.8)5 = 0.32768
 Next, what is P(X=1)?
 One correct answer, followed by 4 wrongs …
hence P(X=1) = 0.2 x 0.84 … Agree?
19
Example: MC Quiz
 P(X=1)
 Answer
 SFFFF (S=success, or correct answer)
 FSFFF (F=failure, or wrong answer)
 FFSFF
 FFFSF
 FFFFS
 The required probability is 5 x 0.2 x 0.8 4 = 0.4096

 Next, what is P(X=2)?

20
 Counting Techniques
Rule of Combinations
 The number of combinations of
selecting X objects out of n objects is:

n n!
C
n X    
 X  X!(n  X)!
 

where n and X are non-negative integers

Defining n! = n(n - 1)(n - 2) . . . (2)(1)
Factorial (!) X! = X(X - 1)(X - 2) . . . (2)(1)
0! = 1 (by definition)

21
 Counting Techniques
Rule of Combinations
 Eg1: An ice cream shop carries 31 possible flavors. How many
combinations of 3-scoop cones are possible?
 Select X=3 from n=31, the number of flavors:
 31 31! 31! 31  30  29  28!
C
31 3   
 3  3!(31  3)! 3!28!  3  2  1  28!  31  5  29  4495
 
 
 Excel command: =combin(31,3)

 Eg2: MC Quiz (n=5)

22
 The Binomial Formula

Eg: MC Quiz
P(X) = Probability of X successes in n trials,
(n=5, π=0.2)
with probability of success π on each
trial X P(X)
0 0.32768
1 0.4096
X = Number of ‘successes’ in sample, 2 0.2048
(X = 0, 1, 2, ..., n) 3 0.0512
4 0.0064
n = Sample size (or number of trials)
5 0.00032
π = Probability of “success”
Q: Find E(X)

23
 The Binomial Formula

Probability of a particular
No. of experimental sequence of trial outcomes
outcomes providing exactly with x successes in n trials
x successes in n trials

24
 Eg: Quality Control
Excel command:
=binomdist(X,n,π,cumulative)

• If the probability of selecting a defective PC is 0.02, find the

probability of selecting 2 defective PCs in a lot of 10

• Find the probability of selecting at most 1 defective PCs in a lot of 10

• What assumption(s) is made in your calculations?

• ANSWER

25
 Properties of the Binomial Dist
 Properties (again)
 Identical trials
 The probability ___ is constant for each observation
 Eg: Probability of getting a tail
 Observations are independent

 More on Independence: There are two sampling methods that

ensure independence between observations
1. Infinite population without replacement
2. Finite population with replacement

26
 Shape of the Binomial Dist
 Eg: MC Quiz Again
 Is it a symmetric distribution?

 The shape depends on the values

of π and n

27
 Shape of the Binomial Dist
 Conditions for Symmetry
 A binomial variable follows a
symmetric distribution if
1. π = 0.5, or
2. Both nπ and n(1-π) > 5
If (2) is satisfied, the binomial
would look like a normal distribution

 Q: Are they symmetric?

1. n=10, π=0.5
2. n=10, π=0.2
3. n=100, π=0.2
 Visualize the answer with simulation (by M. Bognar, University of Iowa)
28
The Binomial Dist Characteristics

 Mean:
μ  E(X)  n

 Variance: σ 2  n (1 -  )

 Standard Deviation: σ nπ (1 - π )

Q: These three formula apply to ____ binomial distributions

(a) symmetric
(b) asymmetric
(c) all

29
 The Binomial Dist Characteristics
Eg1

Shape?

Eg2

30