Chapter VII

PROBABILITY

Approaches
Introduction Definition -Different
Contents: Conditional
Events -

Sample Space
-

Random Experiments
and Multiplication Theorems
Probability - Addition in Economics
Combinations Applications of Probability

7.1 INTRODUCTION

dates back to
the 17h
Development ofthe theory ofprobability
of
century. It originated to explain the pattern possible
outcomes of
certain games of chance such as tossing coins, throwing dice and
drawing cards from a pack of playing cards. In all of these games
outcomes are uncertain. For example, if a coin is tossed either the
head or the tail may turn up. However, we are uncertain about the
actual result until the trial is over. In the same manner if we throw a
die once any one of the six faces i.e., 1, 2,3, 4, 5 or 6 may turn up.
Again, while we draw a card from a pack of playing cards we are
not sure what would be the drawn card. The term probability is always
related to uncertainty.

7.2 DEFINITION

Probability of an event can be defined as the numerical

expression of the likelihood of the occurrence of that event. It 1s
a number which
always lies in between zero and one. Thus if P{A)
is the probability
of event A, then 0<P(A)S 1. IfP(A)= 0, it implies
that the event A will not occur
and if P(A)=1 it means that the
A will
definitely occur. Closer the value of P(A) to unity, higherevent
w
be its chance of
occurence. On the other hand, if the value of P{A) 15
nearer to zero, then lesser
will be its chance of
occurrence.
 185

7.3 DIFFERENT APPROACHES

The different approaches in the theory of probability are:

(a) Classical approach, (b) Relativefrequency theory,

(c) Subjective approach and (d) Axiomatic approach.

a. Classical Approach
Classical approach is otherwise called the mathematical or
deductive (apriori) approach. The basic notion behind this theory is
that the outcomes of a random experiment are equally likely and
mutually exclusive. According to the classical school, probability of
an event A, which is denoted as P(A), is the ratio of the number of
favourable cases to the total number of equally likely and mutually
exclusive outcomes.Thus,

P(A) =
where, m =
number of favourable cases,
n =
total number of outcomes
As an illustration, consider
the experiment of tossing a fair
coin once. Let the occurrence of head is considered
as event A.
Here the total number of outcomes are two
i.e., head and tail. Out of
these two outcomes one outcome
(head) is favourable to the event A.
Then the probability of event A can be
estimated as,
m
P(A)= 2
Probability of the non-occurrence of an event is denoted as
A) or P(A). If 'n' is the total number of
possible outcomes ot
wcn 'm is the number of
favourable cases, then (n m) will be the -

number of unfavourable
cases.
Thus, P(A') =.
n =
=

n
1 P(A) -

Wote: P(A) +
P(A') P(A) =
+1 P(A) =
I}
Illustration
-

that
* the nu -1. A six faced die is thrown
ine once. Find the probability
number on it is (a) (b) greater than 4 (c) even.
5
 188

b. Relative Frequency Theory

In the relative frequency approach, probabilities are obtained
by conducting experiments. Therefore, this approach is sometimes
called the inductive or empirical (a posteriori) approach.
If 'n' is the number of trials in which an event A occurs
'm
times, then relative frequency of the event is the ratio, m/n.
to the relative frequency
According
approach, probability of the event is the
value of relative frequency when the number of
trials,
n tends to
infinity. Symbolically,
Lt(m/n)
P(A) no
where, n = number of trials

m number of times event A has occurred

(Note: Any experiment to collect statistical data is called a
trial)
c. Subjective Approach
The subjective concept of
of an individual
probability measures the confidence
regarding the occurrence of an event. According to
Winkler and Hays, "In the
subjective personal interpretation of
or
probability, probability is
interpreted as a measure of the degree
of belief or as the quantified judgement of a
As this approach lacks particular individual".
objectivity, it is highly flexible and unreliable.
d. Axiomatic Approach

Axiomatic approach to probability is based on

set theory. Three
important concepts, viz., random experiments,
events are used in this sample space
and
approach. 'The basic principles or axioms of
the axiomatic approach are the
following:
1. Probability of the sample space is equal to unity i.e., P(S) =
1.
2. Probability of any event
always lies in between 0 and 1:
i.e., 0 P(A) < 1
3.Iftwo events A and B are mutually
exclusive, then the
of the events A or B is the sum of the probability
individual probabilities of
A and B i.e., P(A U
B) =P(A) +P(B)
 189

7.4 RANDOM EXPERIMENTS

Experiments that can be nepeated under the s a m e conditions

cannot be predicted in any repetition
are
which
and the outcomes of consider the experiment
As an illustration,
calledrandom experiments. two possible
coin repeatedly. In a single toss there are only
oftossing a
head or the tail will turn up. Probability
of
i.e., either the
outcomes tossed for
% in the first toss. Ifthe same coin is
getting head, P(H) = outcomes and
also total number of possible
the second time then without changing
head will not change. Again,
probability of getting can be repeated any
number
the initial conditions the same experiment
it is a random experiment. Throwing
dice, drawing
Therefore,
oftimes. cards with replacement etc., are also
cards from a pack of playing
random experiments.
as a randomm
be considered
Generally, any experiment can
conditions:
experiment if it satisfies the following
1. The experiment must be capable of
repetition.
one outcome for the experiment.
2. There should be more than
in
defined event must remain the same
3. Probability of an already
each repetition.
cannot be predicted.
4. Outcome of the experiment in any repetition
7.5 SAMPLE SPACE
outcomes
universal set which contains all of the possible
The
as its elements is
the sample space of that
f a random experiment
experiment. Usually sample space
is denoted as the universal set, S
The sample
and elements of the samplespace are called sample poinis.
a coin once is, S {H, T}. =

space of the random experiment oftossing

ofthe experiment ofthrowing a die
once
Similarly, the sample space
Can be given as, S = {1,2,3, 4, 5, 6}.
of the elements in the sample
In the above cited examples, all
are called
Space are arranged in a single row. Such sanmple spaces
consider another random
-dimensional sample spaces. Now,
of
Cxperiment of throwing two dice simultaneously. Sample space
this experiment is given as:
 191

A-{I, 3, $} and
S { 1 , 2 , 3, 4, 5,61.Let, the
throwing a die once, also A and B are events
as they are

In this example
R-2,4,6
subscts of S.
it
as its element,
contains only a single sample point
Ifan event two or more sample points
as ther

simple event. Events having

is a
For example. in the random
called compound events.
elements are
I,2,3,4,5, 6}
die once, the sample space, S- {
experimentofthrowing
a
is a simple event
while B Is a
and B {1, 3,51. Here, A
Let A {2
compound event.

Mutually Exclusive Events

if thev
Tiwo A and B an said to he mutuall echusive
events

cannot occur simultaneously In

other words, if the occurrence of
one event in itself prevents
the simultaneous occurrence of another
For example.
event, then both of the events are mutually exelusive.
a card trom a pack of
consider the random experiment of drawing
of a red card and B is the
playing cards. Let R is the event getting
event of getting a black card. Events
R and B are mutually exclusive
as drawn card cannot be a red card and a black card at the same
the
time. Tocite another example, let, S {l,2, 3,4,5,6},A-{ I, 3, 5
and B 2,4,61. Here, A and B are mutually exclusive as they have
no common sample points and hence they cannot ccur sinultaneously

Iftwo eventsA and B are mutually exclusive, A nB will be a mull

set with no elements.

Partially Overlapping Events

Two events A and B are partially overlapping if both ot them

simultaneously. Such event sets are joint sets with at loast

can occur
As an illuistration, consider the rando
one common sample point.
A is the event of gettin
experiment of throwing a die once. Let
event of getting an odd number.
As 3 iN a
number 3 and B is the the two events can
in both of the event sets,
common sample point
hence theyare partially overlapping
events
OCcur simultaneously and

given below mutually exelusive events

In the Venn diagrams
events are shown
and partially overlapping
 192
events
overlapping
Partially
events
(i)
exclusive
S
(i) Mutually
S B
A
B
A

Equaly Likely Events

equally likely
events if the probability
Two events A and B
are

events are
For example, let A is the
same.

of the occurrence of both tail when a fair

B is the event of getting
of getting head and
event events are equally
coin is tossed once. As, P(A) P(B) 2, the two
=
=

likely events.
Events
Dependent and Independent
o c c u r r e n c e or non
Events are said to be dependent if the
the probability of any other
event.
o c c u r r e n c e of one event influences
As an illustration, assume
that in a basket there are 10 good oranges
are taken one after the
other from
and 2 bad oranges. If two oranges the
it without replacement, outcome
of the first trial will affect
outcome ofthesecond trial. Suppose A,
is the eventof getting a good
orangen
in the first trial and Az is the event of getting good
a
orange
the second trial.
m_10 5
P(A)n 12 6
and it is not
If in the first trial we have obtained a good orange
Therelore,
replaced then for the second trial, m 9 andn= 11.
=

P(A) 91
not
and it was
But if in the first trial we have got a bad orange
Thus, u
before the second trial, P(A,) would be 10/11.
replaced
 193

utcome of the first trial intluences the outcome of the second trial.
Hence, the two events A and A, are dependent events. Drawing
cards in succession from a pack of playing
cards without
is another example of dependent events.
replacement
Events are said to be independent if the
occurrence or non-
occurrence of one event does not influence the
occurrence of another event. For
probability of the
example,
consider the random
experiment of repeatedly tossing a coin. Let A, is the event of getting
head in the first toss and A, is the event of
toss. Whatever be the outcome of the
getting head in the second
first trial, it does not affect the
probability A. Hence. A and A, are independent events. Drawing
of
cards one after another from a
pack with replacement after each
trial is another example of
independent events.
Exhaustive Events
Two events A and B are said to be
events if taken together they contain all
collectively exhaustive
the
of possible outcomes of a
random experiment i.e., if (AUB)=S, A and B are exhaustive
events.
For example, if we toss a coin once there are two
possible outcomes,
H and Ti.e., S= (H,T). If event,
exhaustive events as (AUB) = S.
A= {H} event, B {T},
and =

they
are

7.7 CONDITIONAL PROBABILITY

Suppose A and B are two events. Conditional probability of A
given B, which is denoted as P(A/B), is the probability of Awhen it
is known that the event B had already occurred. As an illustration,
consider the random experiment of throwing a die once. Let A is the
event ofgetting number 3 and Bis the event of getting an odd number
Symbolically, A = {3} and B= {1,3,5}. Probability of event A, i.e.
P(A) = 1/6. But if we know that the event B had already occurred,
then the sample space contains only three sample points 1,3 and 5 ie.
n 3. Hence, P(A/B) 1/3. To cite
= another example, consider the
of 52 playing cards. Let D
experiment of drawing a card from pack
a

s the event of getting a diamond

card andR is the event of getting a
that the drawn card
13/52 =
4. However, if we know
red card. P(D)
=

happened, then P(D/R) = 13/26 %.

red i.e., event R had already
 194

INote: I1, A and B are two independent events P(A/B) PCA) and
P1/A)PB)
Conditional probability can beestimatedusingtheformulae
PMA/B) P(AnB)

P(B)
PAB)
P(B/A)
P(A)
INote: PAnB)is the joint probability of the occurrenee ofboth of
the events A and B|
7.8 THEOREMS OF PROBABILITY

Two inportant theorems of probability are:

. Addition theorem and
2. Multiplication theorem
7.8.1 Addition Theorem (Additive Theorem)
Addition theorem of probability has two cases.
Case -1. When events are mutually exclusive
According to addition theorem if A and B are two mutually
exclusive events, then the probability of the occurrenee of A or B,
which is denoted as P(AUB), is the sum ofthe individual probabilities
of A and B. Thus,
P(AUB) = P(A) + P(B)
The above theorem can be extended to any number ofmutually
exclusive events. Thus,
P(AUBU C) = P(A) + P(B) + P(C)

P(AUBUCUD) P(A) + P(B) + P(C) +P(D)

Illustration -4. Adie is thrown once. What is the probability of getumg
number 4 or 6?
g 6.
Solution: Let A is the event of getting 4 and B is the event of getng
As the two events are mutually exclusive.
 195

P(AB) P(A) + P(B)

PMA)- P(B)-
PAB) 2L
6' 6 6b
llustration-5. There are 240 workers in a factory. Out of them 30
Are from Kerala, 20 are from Tamilnadu
and the rest are from Punjab.
Ifa worker is selected at random from that
that he
factory, find the probabil-
ily is a
person either from Kerala or from Tamilnadu?
Solution: 1Let K be the event that the selected
and T be the event that he is from
person is from Kerala
Tamilnadu. As the two events are
mutually exclusive,
P(KUT) - P(K) + P(T)
30 20
PK) n 240
240 PCT)2400
30 20 30+ 20
P(KT) 240 240 240
() 5
240 24

Case -2. When events are not mutually exclusive

Iftwo events A and Bare nol mulually exclusive then addition

theorem is stated as,
P(AU B)- P(A) + P(B) - P(AnB)

[PAB) is the joint probability of events Aand B

INote: If A and B are mutually exclusive events, P(A nB)- 0, as

AOBis a null set with no elements|

student pusscs in a subject A is 0.8
lustration -6. Probability that a

and the probability that he passes

in another subject B is 0.9, What is
the probability that he will pass in
either A or 137

event that the

student passes in subject A and
olution: Let A is the
 197

P(MUK)= P(M) + P(K) P(MnK)

-

P(MUK)=0.5 + 0.4 0.06 0.84

person reads neither Manorama nor Koumudi,

ibability that a
P(MUKy = 1 - P(MUK)

= 1 - 0.84 0.16

Number ofpersons who read neither ofthe papers can be estimatedby

the obtained probability with the total number ofpersons.)
maultiplying
Number of persons who read neither
=
0.16 x 500000 80000

1.8.2 Multiplication Theorem

Like addition theorem, multiplication theorem also has two cases.

Case -1. When events are independent

According to the multiplication theorem, iftwo events A andB

are independent, probability of the occurrence of A and B, which is
denoted as P(AnB), is the product ofthe individual probabilities of A
and B. Thus,
P(ANB) = P(A) x P(B)

This theorem can be extended to any number of independent

vents. For example ifA, B and C are three independent events,
P(AnBnC) = P(A) x P(B) x P(C)

Illustration -9. A coin is tossed twice. Find the probability of getting

head in both trials.
Solution: Let H, is the event of getting head in the tirst toss and H,
second toss. As the two events are
is the event of getting head in the
independent, P(H,nH:)= P(H,) P(H,) *

%
P(H)=%, P(H,)=
P(H, H) =% =% x

Statistics is given
to two students. Their
llustration -10 A problem in What is the probability
and % respectively. y
chances of solving it are %
solvedf"
that the problem is not
 198

Solution: Let A, is the event that the first student solves the nraki.em
and A, is the event that the second student solves it. The
not be solved if both of them fail to solve it.
problem wi
will
Probability that the first student fails to solve it,

PMA= 1-P(A,) =1-=

Probability that the second student fails to solve it,
P(A =
1 -

P(A) =
1
--
As the two events are
independent,
PMA nA) =

P(Ay *
P(A,) = x= 12
Probability that the problem is not
solved = 12
Illustration -11. Three cards are
after replacing the drawn from a
all the three cards
drawn card in each trial. pack of playing cards
Find the
are aces.
Solution: Let Aj, Az and
probability that
second and third draws Az are the events of getting ace in
the first,
the three events respectively.
As the trials are with
are
independent. Hence, replacement
P(A, OA2 nA.) =
P(A,) x
P(A,) P(A,)
*

(A) P(A,)
52 =P(A,)=
52
PCAnAnA)- =
1
Case -2: When 2197
events are
If two
events A and B
dependent
rence of A and are
B is
given as, dependent, probability
probability ofor uthe occur
P(AnB) P(A) P(B/A) =
x
or
P(AB) =P(B)
P(A/B) x
 $005

10 APPLICATONS OF PROBABILITY IN ECONOMICS

In the early years of its origin, probability theory was used to
explain the pattern of outcomes related to certain games of chance.
Now a days, it is widely used in almost all disciplines like Economics
Statistics, Business Management, Physies, Biology and Engineering.
Major uses of the theory of probability in Fconomics are given below
1.The theory of probability is used to estimate the probable outcomes
n many uncertain situations.

2. Methods oftesting economic hypotheses are based on probability

theory. In this way it helps to formulate many economic and
bussiness theories.

3.Probability theory enables entrepreneurs and businessmnen to

take proper economic decisions.
4.It helps econonmic decision makers like producers to assess the
and programmes
possible impacts of their policies
estimation theories are based on probabilitytheory. It
5.Statistical
enables to predict the future changes in cconomie variables such

as demand, supply, income etc., With reasonable level of accuracy

6 Finally, probability theory Can be successtully used in inventor

uanagenent It is also usea quatity cOntrol tests ofproducts

SUMMARY
habality Is defiwd as ue tueTKAl expression ot the likelibhod

evcnts
olevents The tou
tor
approaches to
an

arethe
of
ouCurenee
probability
the classical aeuroch, relative irequeney approach, theory
subjective
and the atic appac
approa
arionatc
h. ln classical thcory
approach
probability
 CHAPTER VIII

PROBABILITY DISTRIBUTIONS
Contents: Introduction Random Variable
bution Binomial Distribution
Probability Distri
-

(Assumptions and Definition)

-

Properties of Binomial Distribution - Binomial Frequencv

Distribution Normal Distribution
Properties of Normal
-

Distribution Standard Normal Distribution

Estimation
of Probabilities Using Standard Normal Table Uses
-

Normal Distribution. - of

8.1 INTRODUCTION

Probabilities can be estimated either

cally. In an empirical probability distribution, empirically mathemati-
or

outcomes of a random
probabilities of the various
relative
experiment are obtained by observing the
frequencies of outcomes while in mathematical method,
probabilities of outcomes are estimated through logical deduction.
Probability distributions based on the deductive method or
probabilities are called theoretical distributions. Binomial expected
and normal distribution are the distribution
two popular theoretical
distributions. While the former deals with probability
the latter is used to discrete random variables,
estimate the probability of continuous variabies.
8.2 RANDOM VARIABLE
If the value of a variable is determined by the outcome or a
random experiment it is
called a random
variable. In other wora
random variable is function defined over the
a an
experiment and generally assumes different sample
values withspace
a acifinite
probability associated with each value.
As an
illustration,
coin once. Let us defineconsider the random experiment or t inga
the random variable, X as the of
nun
 obtained.If the outcome of the trial is a tail, then X takes the
heads

ue 0. On the other hand, if the outcome is head, then the value of

then the random vari-
will be
Xwill 1. Similarly, if a coin is tossed twice,
be 1.
of heads) can
(number take three values i.e., 0, 1 and 2.
able, X then the specific values of X
Generally, ifX is the random variable,
are denoted as Xi, X2, ..., Xn
into discrete and
Random variables can be broadly classified
variable always takes specific values such
continuous variables. If a
variable
as0, 1,2,3, it is called a discrete variable. Acontinuous
. .
,

where a and b are

may take any value within a range. If, a<Xsb,
real numbers, X is a continuous variable.
8.3 PROBABILITY DISTRIBUTION

schedule that shows the various

distribution is a
Probability
associated with each
values of the random variable and probability
value. As an illustration, consider the random experiment tossing
of a

contains four sample

coin twice. The sample space of this experiment
the number of heads turn
points, (H,H); (H,T); (T,H) and (T,T). Let
up is the random variable,
X. The possible values of X may be given as
X 0,=
1 and
X2= 2. Probabilities associated with the various
X3
=

values of X are; P(x) = %, P(x,) = % and P(x3) = %. Probability

distribution of the experiment can be shown as follows:

Probability distribution

No.of heads (X) P(X)

0
1
2

Total

Mustration-1. Write the probability distribution ofthe experiment of

random
throwing a die once by the taking the number turning up as the
variable.
 216

Solution:(Draw a table with two columns. In the first column write the
possible values of X. In the second column list the probability for each
value of X.)
Probabilit distribution
No. turned up (X) P(X)

6
Total

Generally, if xi, x2, ... X are the specific values of a random

variable, X and P(x,), P(x,),... P(Cx,) are the respective probabilities.
then the probability distribution is given as,
X P(X)
XI
P(x1)
X2 P(x2)

Xn
P(xn)
Total
As theprobability of the sample space, P(S) =
1, sum total o
probabilities probability distribution, i.e.
in a

P(x)+ P(x») +P(x«) =1

+

8.4 BINOMIAL DISTRIBUTION

Binomial Distribution was first
introduced by James Bermou"
to deal with random
experiments having two categories ofoutconme
viz., success and failure. Mathematical model of distribution
binomial
 217

s hased on the folowing assumptions:

1.The experiment is repeated under similar conditions

for a speci-
fied number of times, n.
2. In each trial there are only two categories of outcomes
success and failure. namely,
3.Probability of success, which is denoted as 'p' remains the
ineach trial. Probability of failure is denoted as same
and q=1 p. q also
remains the same in
'q'.As, p+q=1
4. All the n trials are each trial.
any one trnal does
independent. It means that the outcome of
not affect the outcome of
any other trial.
Binomial probability function is
given as,
px)-C,pq*
where, n number of trials, x=0, 1, 2,. n)
P probability of success
qprobability of failure (1-p),
x= number of success
[Note: n and p are the two
parameters of binomial
The entire binomial distribution
following table. probability distribution is shown in
the
No. of success
Prob. of success
(X) P(X)
0
C,pq-=q
C, p' q-l=npq"-i
C,p q2

C.p"qP
Total
 218

Nore: ",C,- l: p"= 1.f=1. Probability ofzero success, ie.,

0)= q and probability of n success, p{x =n)= p"]
This distribution is called binomial because the probabilities
specitied above as, q, npq", "C,p*q**, . P "are the successive
terms in the binomial expansion
of (q+ p.
lllustration-2. Ifx is a binomial variate with parameters 10
what is the probability that x takes the value zero?
and 0.3,
Solution: Given, n= 10,p=0.3
q1-p=1-0.3 =0.7
Probability when x is zero,
P(x= 0) = q = 0.710

Illustration -3. Three fair coins are tossed once. If

occurrence of
head is considered as success, write down the
binomial probability
distribution of the experiment.
Solution: (In this experiment
possible values of x are 0, 1,2 and 3)
Binomial probability function is
given as,
p(x)=C,p'q (x0,1,2,.n)
Given, n= 3, p=%,q=1-%=%,the probability distribution
of the experiment can be given as,
No. of heads
(X) P(x)
q-(4) =%
npqt=3 x hx (4P=%|

C,pq-2=3x (4} =%
3
p= (=%
Total
llustration -4. An unbiased coin is tossed five times. Find the
probability of getting (a) exactly two heads (b) atleast 1our head
(c) atmost two heads.
 222

8.5 PROPERTIES OF BINOMIAL DISTRIBUTiON

The binomial probability function is given as,
p(x)= "C,p'q"* (x =0, 1, 2, .., n)
This distribution is widely used to estimate probabilities ofevents
when the outcomes of an experiment can be classified into success
and failure. Important
properties of the distribution are given below:
1. Arithmetic mean of binomial distribution is the of product the
parameters n and p. i.e., u= np
2. Standard deviation of the distribution is
given as,
G Vnppq
Variance, o' = npq

3. First four central moments

of the distribution are estimated as,
i)4 =0
(i) u,= npq
(ii) u,= npq(q - p)
(iv) = 3n°p?q + npq(1 6pq) -

4. Skewness, Vß, = -9-p)

npq
[Note: If p <0.5, the distribution is positively skewed and if,
it is p> 0.5
negatively skewed.]
5. Kurtosis,B==3+Ópg
npq
6. In a binomial distribution mode is the
it is not an
integral part of (n + 1}pir
integer value. If (n +1)p is an integer there are two
modes. They are, (n+1)p and
{(n +1)p} -1.
For example; if p =
0.5 and n =
20, (n +1)p =
10.5
Mode, Mo =10
If, p =0.25 and n= 27, (n + 1)p 28 x 0.25 7 =

As, 7 is an integer there are two modes. They are 7 ana o.

 223

s6 BINOMIAL FREQUENCY DISTRIBUTION

Let us assume that 'n' trials constitute an experiment and the
experiment is repeated 'N' times. For example, if 5 coins are tossed
20 times. n =5 and N 20. The frequency function of
=

binomial
distribution is given as, f(x) N x p(x).
=

As. plx)=C,pq*

x)= Nx C, p'q**
As an illustration, consider the random experiment of tossing2
coins 20 times. Here, n2
and N=20. Binomial frequency distribution
of the experiment can be written as follows:
No. of heads (X) P(X) =
"C, p'q*fx) =NxP(x)
0 20x = 5
1 20x = 10
2 20 x = 5

Total
20

Similarly, if 5 coins are tossed 320 times, n =

5 and N =
320.
Frequency distribution of this experiment is given below:

No. of heads (X) P(X) f(x) =

N x
P(x)
32 320x /32 =10
/32 320x /32 =50
19/32 320x /32 = 100
3 10/32 320x 0/32 =100
532 320x /32 = 50
5 320x /32 =10

Total 320
 226

Binomial probability function, p(x) ="C, p°q"* (x =

0,
(a) Probability of having exactly one boy,
1,...n)
p(x = 1)= npq"

= 4x 0.5 (0.5) x
=
0.25
Number of families having exactly one boy = Nxp(x =1)

500x 0.25 125

(b) Probability of having exactly two girls is same as the probability
of having exactly two boys as, n = 4.

P(x = 2)="C,p?qn-2 =iC,(APA}

= 6x 0.0625 =0.375
Number offamilies having exactly two girls = N x p(x=2)

500 x 0.375 188

8.7 NORMAL DISTRIBUTION

The most popular theoretical probability distribution used to deal

with continuous variables is normal distribution. It was first introduced
by Abraham De Moivre as a limiting form of binomial distribution.
Binomial distribution becomes a normal distribution under the folowing
situations:
I. when the probability of succe_s, p and probability of failure, q
are more or less equal i.e., p*q

2. when the number of trials, n is very large

In the limiting case when, n -> oo, even if p and q are not equal
or nearly equal, binomial frequency curve takes the shape of a normal
curve.

A Continuous random variable, x is said to be normally distrib

uted ifit satisfies the following probability density function,

-(x-H
e 20
P 21
aV 27T
(-00x o )
 227

where. e 2.7183 (base of natural logarithm)

T 3.1416
H arithmetic mean of the distribution
G =
standard deviation of the distribution
In the normal probability density function, e and n are
the two
Onstants while ji and o are the two parameters.
A normal distribution with u as its
arithmetic mean and o- as
its variance is usually denoted as,
P(x)- N(u, o)
Note: The above function is read as,
tion with parameters u and o)
p(x) follows a normal distribu
Total area under a normal curve
As normal distribution is a represents total probability.
probability density
ie., p(- oo Sxs»)is equal to unity. Hence,function,
total probability
the total area under the
normal curve is taken as 1. Area under the
two points 'a' and 'b' shows
normal curve in between
the probability of x in
between 'a' and
b'.Symbolically, p(a<x<b) is the area
under the curve in between
the two values 'a' and 'b'.
In the following
diagram the shaded area between the
a' and 'b' represents the points
and 'b' i.e., p(a < x S b).
probability of x lying between the values*a

Normal Curve

p(a x b)

b X

8.8 PROPERTIES OF NORMAL DISTRIBUTION

Important properties of normal distribution are the following:
Normal curve is bell shaped. It implies that ifa perpendicular is
drawn to the base line through the point x = u, it will divide the
 228

otal area under the curve into two equal parts i.e..

S S p) p(u SxS)05
2All od oentral moments ofthe distribution are equal to zero
1
A , 0 The second central moment
fariance) and fourth central moment. u,3g'
3Mean, median and mode of a normal distribution coincidei
Mean Median Mode
4 Normal distribution is a uni-modal distribution.

5.The maximum ordinate value (height) of the normal curve is

av 2
6.Area under the normal curve is distributed as follows:
t o covers 68.279% of the total area
H 20 covers 95.45% of
the total area and
H 3o covers 99.73%
of the total area
The
distribution of area under the normal curve is
matically shown below diagram

P(x) N(H, O)
68.27 %
95.45 %
99.73 %
-3o -20 H-g
7.Normal curve is asymptotie to the X axis. It means tna when
the value respective tails ol
of x moves towards and o, the respective a
-oo
Uhe curve tend to be closer and closer to the base, Dut
touch it.
8. The two points of intlection of a normal at the
Curve o c c u r
onts

H + and x =u- a. h Cun atir ture

These are the points where u

 229

of the curve) cnanges rom convex to concave and from

inature
natu
concave to c o n v e x .

Nomal curve is a continuous curve from o too. However, as

-

99.73 percent of the area under the curve lies in between u + 3o

to 3o.
and - 3o, usually the range otf the curve is taken as -3o
distribution. Hence, skewness
10. Normal distribution is a symmetric
ofthe normal curve is zero. i.e.,
Skewness, vB,- 0

1. Normal curve is mesokurtic i.e.,

Kurtosis, B,- 3
distribution lie at equal
12. The first and the third quartiles ofnormal
distance from median i.e.,
M-Q =Q M

13. Quartile deviation and

mean deviation ofthe distribution are
given as,
QD 0.6745 o or
o approx.)
(approx.)
MD 0.7979 o or

NORMAL DISTRIBUTION
STANDARD
8.9
of normal
have seen, u and o are the two parameters
As we
a will be different and
data, values of ju and
distribution. For different form af
c u r v e s also
will be ditfferent. A standardised
normal
Iscalled standardnormal
therefore =
I
and
distribution by taking =0 o
normal variable (variate) is usually denoted
While the normal
distribution.
normal variable
is denoted as z
standard
as x, the
and SD=o
=
can be converted
curve with AM
A normal changing both origin
and scale. The
curve by norma
into a standard normal variable, X into the standard nal
normal
convert the
formula used to
as,
variable, z is given x-H
z
 230

The standard normal probability function is given as,

P)-
V2T
INote: To parameters of standard normal distribution ane 0 and n

The standard normal curve is a bell shaped

continuous eurva
that ranges from -co to oo. The maximum
point the curve occurs at
ot
Z=0 as shown in the
following diagram.

p(z)- N(O,1)

-C0
0

8.10 ESTIMATION OF PROBABILITIES USING

NORMAIL TABLE
STANDARDD
As standard normal distribution is a
probability density func-
tion, total area under the standard normal curve is
p(-o S z so) 1. Again, the probability that theequal unity i.e.
to
=

standard normal
variable, z lies in between two
points 'a' and "b' is the area under the
curve in between 'a' and "b°'. In
the following
z
lying in between 'a' and "b' is shown as thediagram, probability of
shaded area below the
standard normal curve.

p(z) N(0,1)

P(a s z sb)

a b