CHAPTER

2
• Conditions of statistical dependence and independence,

• Bayes‘ theorem and its applications. Probability distributions

– Random variable,

• Expected value, Binomial, Poisson, Normal and Exponential

distributions

1
Objective :

• Use of Probability theory in decision making

• Rules to calculate probability

• Definition and use of Bayes theorem

• Probability Distributions

2
Probability

• Jacob Bernoulli, Abraham de Moivre, Thomas Bayes, Joseph Lagrange

are the pioneers in developing the Probability techniques.

• Theory of probability is applied to social and economic theories,

insurance industry required precise knowledge about the risk of loss
in order to calculate premium, decision making research.

3
Probability
• In how many ways can a cricket team choose a captain and vice captain from the
team ?

Answer : There are 11 members in a cricket team, a captain an be chosen in 11

ways and the vice captain in 10 ways.
Thus choosing a captain and vice captain can be done in 10 * 11 = 110 ways.

• There are 3 roads connecting city A to city B and there are 2 roads connecting city
B to city C then in how many ways can a person travel from city A to City C ?

Travelling from city A to city B is done in 3 ways and from city B to city C in 2 ways.
Hence total number of ways in which one person can travel from city A to city C are
3 * 2 = 6 ways.
4
Probability

Terms used :

An event : When you toss a coin getting a tail would be an event and
getting a head would be another event.

An experiment or trial : The activity which can be repeated and has

well defined set of possible outcomes.

The set of all possible outcomes is called as Sample space.

5
Probability
Statistical dependence and independence
Independent event is an event that has no connection to the
happening of another event.

Dependent event – one event influences the probability of another

event.

Eg Tossing a coin, rolling a die

Parking and getting a parking ticket

Survey a group of people to analyse the product effectiveness
6
Probability

Probability is the measure that an event is likely to happen.

So either it is 0 or 1, 0 indicates impossible events and 1 possible.
Higher the probability more likely the event will occur.

When tossing a coin both outcomes are probable. Thus probability of

head or tail is 0.5 or ½.

3 ways to classify probability :

7
Probability
1. When you toss a coin :
a. Getting a head or tail is an event
b. This activity or procedure is called an experiment
c. The sample space is { head, tail }

2. Tossing a die : Sample space = {1,2,3,4,5,6}

E = {2,4,6} is an event. ie getting even number.

Toss 2 coins – HH, HT, TH, TT

Example – Give sample space : Give list of possible outcomes of tossing 2 dice
3 coins are tossed simultaneously
A committee of 2 is to be formed out of persons a,b,c,d

Mutually exclusive event : are things that cannot happen simultaneously.

8
36 cards
16 cards

9
Exercise

• In a survey conducted at a restaurant 20 people preferred orange

juice, 16 preferred grapefruit juice and 9 preferred apple juice with
breakfast. If a person is selected at random find the probability that
he or she prefers grapefruit juice ?

10
Probability : Exercise

Mutually exclusive event :

• 2 events cannot happen at the same time
• Eg tossing a coin

Examine following events are mutually exclusive in drawing a card from deck of
cards :
1. A heart and a queen

2. A club and a red card

3. An even number and a spade

4. An ace and an even number 11

Probability
There are 3 types of Probabilities :

1. Classical approach :

Probability of an event = number of outcomes where the event occurs

total number of outcomes

Probability of getting head = 1 ( No. of outcomes of 1 toss)

1 + 1 (Total number of outcomes)

Eg. Rolling a 2 on a die = 1/6 (possible outcomes 1,2,3,4,5,6)

Choosing a card from a deck of cards = 1/52
12
Probability
2. Relative frequency of occurence :

• This method uses frequencies of past occurrences as probabilities.

• We determine how often something has happened in the past and use that figure
to predict the probability that it will happen again in the future.

• Suppose an insurance company knows from past data that all patients above 40
years old , about 60 out of every 100000 will die within 1 year period. The
probability of death for that age group is
60 / 100000 = 0.0006

Relative frequency = No of successful trials / Total number of trials.

Large number of trials are required.

13
Probability

3. Subjective Probabilities :

These are based on the beliefs of a person

making the probability assessment. This is the probability assigned to
an event by an individual based on evidence available.

A judge is deciding whether to allow the construction of a nuclear site

where there is some evidence of a fault. He must use his best
judgement to find out the subjective probability.

14
15
Probability : Exercise
The HR has drafted a set of wage and benefit demands to be presented to the
Management. To get an idea of worker support he randomly takes feedback from
Machinist (M) and inspectors (I). He polls 30 with the following results.
Opinion of Package M I
Strongly support 9 10
Mildly support 11 3
Undecided 2 2
Mildly oppose 4 8
Strongly oppose 4 7
30 30

1. What is the probability that a machinist selected at random mildly supports the
package ?
2. What is the probability that an inspector is undecided about the package ?
3. What is the probability that a worker (M or I) strongly or mildly supports the
package ? 16
Probability : Exercise

1. Determine probabilities of the following events in drawing a card from a

standard deck of 52 cards.

1. A seven

2. A black card

3. A red face card ( king , queen or jack )

4. Ace

17
Probability Rules
Venn Diagram

A B A B

Two Mutually Exclusive event Two non mutually exclusive events

It is visual representation of relationship between sets, subsets or units.

Addition rule : If 2 events are not mutually exclusive it is possible that both will occur. Eg. What is
the probability of drawing an ace or heart from a deck of cards ? The event can occur together thus
it is not mutually exclusive.
P(A or B) = P(A) + P(B) – P(AB)
P(A or B happening when A and B are not mutually exclusive)=P(A happening) + P(B happening) –
P(A and B happening together)
P(drawing ace or heart)=4/52 + 13/52 – 1/52 = 4/13
18
Probability Rules

• Probability (A or B) = P(A) + P(B) for Mutually Exclusive events.

Example : There are 4 candidates selected for a job interview. What is

the probability that 1st will be selected. P(1) = ¼
• What is the probability that second will be selected ? P(2)=1/4
• What is the either first or second will be selected ?
• P(A or B) = ¼ + ¼
• Above are Mutually Exclusive events.

19
Probability Rules : Example
• 23
A B
6
8 13

• From the diagram which indicates number of outcomes of an experiment

corresponding to each event and the outcomes not related to either events. No
of outcomes = 50
• Find P(A), P(B) and P(A or B).
P(A) = 14/50 = 0.28 P(B) = 19/50 = 0.38
P(A OR B) = P(A) + P(B) – P(AB)
= 14/50 + 19/50 – 6/50
= 27/50
= 0.54
20
Problems
• A bag contains 30 balls numbered from 1 to 30. One ball is drawn at
random. Find the probability that the number will be multiple of
a. 5 or 9 b. 5 or 6

• One card is drawn from a pack of cards. What is the probability that it
is either a king or a queen ?

21
Probabilities under conditions of statistical
independence
The occurrence of one event has no effect on the probability of the
occurrence of any other event.

There are 3 types :

Marginal

Joint

Conditional
22
Marginal or unconditional Probability
• A single probability means only one event can take place. It is called
marginal or unconditional probability.

• When we toss a coin the outcome is 50% chance of getting head or

tail. P(H) = 0.5 and P(T) = 0.5

• If the coin is biased or unfair then the head may occur 0.90 of the
time and tail 0.10 of the time.

• P(H) = 0.90 P(T) = 0.10

• The outcomes of several tosses of this coin is unrelated to the

outcomes that may precede or follow. 23
Joint Probability

The probability of 2 or more independent events occurring together or

in succession is given as product of their marginal probabilities :

P(AB) = P(A) * P(B)

Probability (events A and B occurring together) =

Probability (event A occurring) * Probability (event B occurring)

24
Joint Probability

Example :

The probability of 2 heads on 2 successive tosses is the probability of

head on first toss (H1) and probability of head on second toss (H2)

P(H1H2) = P(H1) * P(H2) = 0.5 * 0.5 = 0.25

Similarily probability of getting 3 heads is

0.5 * 0.5 * 0.5 = 0.125

25
Conditional Probability
We have considered 2 types probabilities, marginal (symbolically
P(A) and joint (symbolically P(AB)).

Symbolically conditional probability P(B/A) ie probability of

occurrence of B given that A already has occurred.

P(B/A) is the probability that the second event B will occur if a first
event A has already happened.

Eg. Given that you drew red card what is the probability that it is 4
?
P(4/red) = 2/26 = 1/13
26
Exercise
• A Grocery store is reviewing its restocking policies and has analyzed the
number of container of orange juice sold each day for the past month
No sold Morning Afternoon Evening
0-19 3 8 2
20-39 3 4 3
40-59 12 6 4
60-79 4 9 9
80-99 5 3 6
100 or more 3 0 6
Total 30 30 30

• What is the probability that on a randomly selected day the number of

cartons sold in evening is between 80 and 99
• Probability that 39 or fewer cartons were sold during afternoon
• Probability that either 0-19 or 100 or more cartons were sold in morning 27
Exercise
Fragrance Soap ltd is a leading soap manufacturer. Active is a well
known brand, Company conducted a survey to find out preference for
the brand. If the customer is selected at random
Prefer Ahmedabad Gwalior Raipur Lucknow
Yes 55 40 80 75
No 40 30 20 90
No Opinion 5 10 20 35

a. What is the probability that he / she prefers active

b. Consumer prefers Active and is from Ahmedabad
c. From Raipur and no opinion

28
Probability Formulas
1. Probability of an event = number of outcomes where the event occurs
total number of outcomes
2. P(A) = Probability of event A happening
3. P(A or B) = Probability of either A or b happening
4. When events are not mutually exclusive : addition rule applies
P(A or B) = P(A) + P(B) – P(AB)
5. When events are mutually exclusive : addition rule applies
P(AB) = P(A) + P(B)
6. The probability of 2 or more independent events occurring together or in
succession P(AB) = P(A) * P(B)
7. Conditional Probabilities P(B/A) = P(BA) / P(A) P(A/B) = P(AB) / P(B)

29
Bayes Theorem
• Bayes Theorem named after Thomas Bayes ( English philosopher ) is one of the
famous probability applications.

• Bayes theorem offers a powerful statistical method of evaluating new information

and revising prior estimates. The new probabilities are called posterior
probabilities.

• Bayes theorem describes the probability of an event based on the prior

knowledge of conditions that may be related to the event.

• P(B/A) = P(BA) / P(A) P(A/B) = P(AB) / P(B)

• When there is an event which intersects both A and B. This event = C.

• Then P(C) = P(B/A) + P(A/B)
30
Bayes Theorem
Assume factory has 2 machines. Past records show m/c1 produces 30% of the output, m/c 2
produces 70% of the items. 5% of the items produced by m/c 1 are defective and 1% for m/c 2. If a
defective item is drawn what is the probability that it was produced by m/c 1 or 2?
A=event of drawing an item produced by m/c1
B=event of drawing an item produced by m/c 2
C=event of drawing a defective item produced either by m/c 1 or m/c 2
P(A) = 0.3 P(B) = 0.7 P(C/A) = 0.05 P(C/B) = 0.01

EVENT PRIOR PROB CONDITIONAL JOINT PROB REVISED PROB

PROB
A 0.3 P(C/A) =0.05 P(AC)= 0.015 0.015/0.022=
0.682
B 0.7 P(C/B) = 0.01 P(BC) =0.007 0.007/0.022=
0.318
P(C)=0.022 1.000

The probability that the defective item was produced by m/c 1 is 68.2% and m/c 2 is 31.8%.
31
Bayes Theorem : Example
The probability of 3 events A,B and C occuring are P(A)=0.35, P(B)=0.45 and P(C)=0.2.
Assuming A,B or C has occurred the probability of another event X occuring are
P(X/A) = 0.8, P(X/B) = 0.65 and P(X/C)=0.3.
Find P(A/X), P(B/X) and P(C/X).

P(A/X) = 0.4427
P(B/X) = 0.4625
P(C/X) = 0.0949

32
Bayes Theorem : Example
The probability of 3 events A,B and C occuring are P(A)=0.35, P(B)=0.45 and P(C)=0.2.
Assuming A,B or C has occurred the probability of another event X occuring are
P(X/A) = 0.8, P(X/B) = 0.65 and P(X/C)=0.3.
Find P(A/X), P(B/X) and P(C/X). P(X/EVENT)= P(EVENT/X)=P(X&EVENT)/
P(X&EVENT)/P(EVENT) P(X)

Event P(Event) P(X/Event) P(X and Event) P(Event/X)

A 0.35 0.8 0.35*0.8=0.28 0.28/0.6325
B 0.45 0.65 0.2925 0.2925/0.6325
C 0.2 0.3 0.06 0.06/0.6325
TOTAL P(X) = 0.6325

P(A/X) = 0.4427
P(B/X) = 0.4625
P(C/X) = 0.0949

33
Bayes Theorem : Example
The probabilities of 3 events : A,B and C are P(A) = 0.35, P(B) = 0.45, P(C) = 0.2.
Assume that A,B,C have occurred the probabilities of another event X occuring are
P(X/A) = 0.8, P(X/B) = 0.65, P(X/C) = 0.3. Find P(A/X), P(B/X), P(C/X)

Event P(Event) P(X/Event) P(X and P(Event/X)

Event)

A 0.35 0.8 0.2800 0.2800/0.6325=

0.4427
B 0.45 0.65 0.2925 0.2925/0.6325=
0.4625
C 0.2 0.3 0.0600 0.0600/0.6325=
0.0949
TOTAL P(X) 0.6325

34
Probability Distribution
• Just like the frequency distributions probability distribution expresses
distribution of probability values over different values.
• Probability distribution is a table or equation that links each outcome
of a statistical experiment with its probability of occurrence.
• Example : Probability Distribution of the possible number of tails
from 2 tosses of a fair coin.

1st 2nd No of Probability of the outcomes

Toss Toss tails
T T 2 0.5*0.5 = 0.25
T H 1 0.5*0.5 = 0.25
H H 0 0.5*0.5 = 0.25
H T 1 0.5*0.5 = 0.25

35
Probability Distribution : Example
• In a class of 100 students, 80 students passed in all subjects, 10 failed in 1
subject, 7 failed in 2 subjects and 3 failed in 3 subjects. Find the probability
distribution for number of subjects a student from the class has failed in.

P(failing in zero subject) = P(x=0) = 0.8

P(failing in one subject) = P(x=1) = 0.1
P(failing in two subject) = P(x=2) = 0.07
P(failing in three subject) = P(x=3) = 0.03

The probability distribution table looks like :

x 0 1 2 3
P(x) 0.8 0.1 0.07 0.03
36
Probability Distribution : Problem

• Construct a probability distribution table based on the following frequency

distribution. Compute expected value of the outcome.

Outcome 102 105 108 111 114 117

Frequency 10 20 45 15 20 15

37
Probability Distribution
Example : construct a probability distribution based on the following
frequency distribution

Outcome Frequency P(Outcome) Outcome *

P(outcome)
125 10 0.08 (10/125) 10 (125*0.08)
105 20 0.16 16.8
108 45 0.36 38.88
111 15 0.12 13.32
114 20 0.16 18.24
117 15 0.12 14.04
Total 125 1.00 111.28
Exp.Outcome

38
Probability Distribution : Problem
• An investor who frequently invests in stock market is examining the possibility of
investing in ABC Ltd. He has broken the potential results of investment into 5
possible outcomes with accompanying probabilities. The outcomes are annual
rates of return on a single share of stock that currently costs Rs.150. Find the
expected value of return for investing in a single share of ABC Ltd. If he purchases
stock whenever the expected rate of return > 10% will he purchase the stock
according to this data ?

Return on investment Probability

0.00 0.20
10.00 0.25
15.00 0.30
25.00 0.15
50.00 0.10 39
Probability Distribution

Variable definition : It is any characteristic, number or quantity that

can be measured or counted.

Discrete Probability Distribution :

The variable can take only restricted values. The month which you are
born can take values from 1 to 12.

Continous Probability Distribution :

On the other hand variable can take on any value within a given range
(including integers).
40
Random Variables
• A random variable, usually written as X, is a variable whose values are
numerical outcomes of a random phenomenon.

• A random variable can be discrete or Continous.

• A random variable is a variable whose values are determined by chance.

• If a random variable is allowed to take on limited number of values it is

discrete variable whereas if it is allowed to assume any number of values
within a given range it is Continous random variable.

• Suppose we observe number of patients visiting a clinic which will range

from 100 – 600 per day which is discrete variable. The patients are
screened for their BP which is a continuous random variable.
41
Determine whether Discrete / Continuous variable
1. The speed of race car

2. The number of cups of coffee that a fast food restaurant serves each day

3. The number of people who play state lottery each day.

4. The weight of elephants

5. The time it takes to complete an exercise session.

6. The sugar levels of all patients admitted to a hospital on a specific day.

42
Expected value of a Random Variable
• Expected value is the fundamental idea in the study of probability
distribution.

• Initially it was widely used by the insurance industry and then later on
adapted by all to make decisions under uncertainty.

• The formula for expected value is the probability of an event

multiplied by the amount of times the event takes place = P(X)*n

• Eg if you toss a coin 10 times the probability of getting a head in each

trial is ½ so the expected value is (the number of heads that you can
expect in 10 trials) = 0.5*10 = 5
43
Binomial Distribution

Widely used probability distribution of a discrete random variable is

the binomial distribution. The binomial distribution describes discrete
and not Continous data resulting from an experiment / process called
as Bernoulli process.

The tossing of coin fixed number of times is a Bernoulli process and the
outcomes of such tossed is the binomial probability distribution.

Formula is :

44
Binomial Distribution
Probability of r successes in n trials : ( Symbol ! Is factorial )
3! = 3 * 2 * 1

45
Binomial Distribution
Probability of 2 successes in 3 trials = 0.375 probability of getting 2

heads on 3 tosses of a fair coin

3!
* (0.5^2) (0.5^ )
1
2! * (3-2)!

Mean of a binomial distribution = np

n = no. of trials p = probability of success

Std deviation = sqr root (npq)

46
Binomial Distribution
• A survey found that 1 out of every 5 citizens have visited a doctor in any given
month. If 10 people are selected at random find the probability that exactly 3
have visited the doctor.

n = 10, p = 1/5 , q= 4/5, r = 3.

P(3) = 10! / (10-3)! * 3! (1/5)3 (4/5)7 = 0.201

• Suppose a die is tossed 5 times. What is the probability of getting exactly 2 fours
?
Probability of success on single trial = 1/6 = 0.167
5!/2!*3! (0.167)2 * (0.833)3 = 0.1612
47
Binomial Distribution : Problem
1. Find the probability with n = 12, p=0.45 and r = 8.

2 The latest nationwide political poll indicates that for Indians who are randomly selected
probability that they are with alliance ABC is 0.55, for party PQR is 0.30 and for party XYZ is 0.15.
Assuming that these probabilities are accurate answer the following pertaining to a randomly
chosen group of 10 Indians ?
a. Probability that 4 are with party PQR
b. Probability that none are with party ABC
c. Probability that 2 are with party XYZ

3 Find the mean and standard deviation :

1. n=16 and p=0.40
2. n=10 and p=0.75

48
Poisson Distribution
• It is probability distribution for discrete data.

• It is named after Simeon Poisson

• In Poisson Distribution we look at only the occurrence of an event.

• Like Binomial distribution we do not look at non-happening or failure of

the event.

• Formula is P(x occurrences). =

49
Poisson Distribution

• Poisson distribution describes discrete data resulting from a process,

occurrences range from zero to infinity in a given range.
• Number of calls received by a call center agent
• Number of accidents at a signal

• This is applied for rare events like plane crashes, car accidents. Widely
used in insurance industry and operations.

50
Poisson Distribution : Problem

1. Suppose we are investigating safety of a dangerous road

intersection. Past records indicate mean of 5 accidents / month.
Calculate the probability in any month of 0,1,2 accidents.(
e=2.71828)

2. The mean number of defective products in a factory in one day is

21. What is the probability that in a given day there are exactly 12
defective products ?

51
Normal Distribution
• It is one of the important continuous data Probability distributions

• Most production processes follow normal distribution

• Several mathematician were involved in development of its theory

• Major contribution from Karl Gauss, 18th century mathematician, astronomer

• The curve has a single peak, unimodal

• For a standard normal curve mean, median, mode are same

52
Properties of Normal Distribution

The normal curve has a single peak.

The mean of the normal distribution lies at the

center.

Because of the symmetry median and mode are also

at the center.

The 2 tails extend horizontally and never touch the

axis.

The area below the curve is one. 53

Normal Distribution

54
Normal Distribution

Z = number of standard deviations from the mean

x = value of the random variable

𝜇 = mean of the distribution

𝞂 = standard deviation of the distribution

55
Normal Distribution
1. Find the area under the normal distribution between z = 0 and z = 2.34.

Table value gives =0.4904 or 49.04%

2. Find the area between z=0 and z=-1.75

The table does not give the –ve values.

Since the normal distribution is
symmetric about the mean the area to
the left is same as area to the right.
Table value = 0.4599

56
Normal Distribution
3. Find the area to the right of z = 1.11.

The area between z=0 and z=1.11 is 0.3665

And to the right obtained by subtracting 0.3665 from 0.5

4. Find the area between z = 2.00 and z = 2.47

z=2 from the table 0.4772

z=2.47 from the table 0.4932
0.4932-0.4772 = 0.0160

57
Normal Distribution
5. Find the area between z = 1.68 and z = -1.37
z = 0 and z=1.68 from the table 0.4535
z=0 and z=-1.37 from the table 0.4147
Total area = 0.4535 + 0.4147
= 0.8682

6. Find the area to the left of z=1.99

7. Find the area between z = 2.00 and z = 2.8

58
Normal Distribution : Problem
Example : There is a training programme to upgrade the supervisory skills. A study
of past participants indicates that the mean length of time spent on the
programme is 500 hours and this is normally distributed with standard deviation of
100 hours.

1. What Is the probability that a participant selected at random will required

more than 500 hours to complete the program ?
2. What is the probability that a candidate selected at random will take between
500 and 650 hours to complete the programme?
3. What is the probability that a candidate selected at random will take more than
700 hours to complete the programme?
4. What is the probability that a candidate selected at random will require less
than 580 hours to complete the programme?
5. What is the probability that a candidate chosen at random shall require
between 550 and 650 hours. 59
60
Normal Distribution : Problem
• The customer accounts at a department store have an average
balance of Rs.480 and standard deviation of Rs.160. Assuming that
the account balances are normally distributed

• What proportion of accounts is over Rs.600 ?

• What proportion of accounts is between Rs.400 and Rs.600?
• What proportion of accounts is between Rs.240 and Rs.360?

61
Excercise

Which probability distribution is most likely the appropriate one to use

in the following cases : binomial, poisson or normal

1. The life span of a female born in 1977

2. The number of autos passing through a tollbooth

3. The number of defective radios in a lot of 100

4. The water level in a reservoir

62
Excercise
• XYZ is the supervisor for a Hydroelectric plant. He knows that the dam’s turbines generate
electricity at the peak when at least 1,000,000 gallons pass through the dam each day. The daily
flow is normally distributed with the mean equal to the previous day’s flow and a standard
deviation of 200,000 gallons. Yesterday 850,000 gallons flowed through the dam. What is the
probability that the turbines will generate at peak rate today ?

• XYZ company has received a big order to produce electric motors for a manufacturing company.
In order to fit in its bearing the drive shaft of the motor must have a diameter of 5.1+0.05 / 5.1-
0.05 (inc). The company’s purchasing agent realizes that there is a large stock of steel rods in
inventroy with a mean diameter of 5.07 and a standard deviation of 0.07. What is the probability
of a steel rod from inventory fitting the bearing ?

• ABC an editor of a publishing company calculates that it requires 11 months on average to

complete the publication process from manuscript to finished book with a std deviation of 2.4
months. He believes that normal distrn will be best to describe the distribution of publication
times. Out of 19 books he will handle this year how many will be complete the process in less
than a year ?
63