Chapter

9 Random Variables and

Mathematical Expectation

Siméon-Denis Poisson, a French mathematician known for

his work on definite integrals, electromagnetic theory, and
probability.

Poisson’s research was mainly on Probability. Poisson

distribution was one of the important invention of Poisson.
Poisson contributed to the law of large numbers and it was
useful for the approximation of binomial distribution. Poisson
distribution is now fundamental in the analysis of problems
concerning radioactivity, traffic, and the random occurrence of
events in time or space.
Siméon-Denis Poisson
(21 June, 1781 – 25 April, 1840) ‘Everything existing in the universe is the fruit of Chance’.

- Democritus

Learning Objectives

� Knows the relevance of Random Variables.

� Understands the types of Random Variables.

� Differentiates the various Types of probability functions.

� Enumerates the functions and properties of Cumulative Distribution.

� Understands the Concept of expectation.

� Applies the Theorems on Expectations.

� Identifies Moment generating functions and Characteristic Functions

Introduction
We have studied the elementary probability in the previous chapter. As an application
of probability, there are two more concepts namely random variables and probability
distributions. Before seeing the definition of probability distribution, random variable
needs to be explained. It has been a general notion that if an experiment is repeated under
identical conditions, values of the variable so obtained would be similar. However, there
are situations where these observations vary even though the experiment is repeated under
identical conditions. As the result, the outcomes of the variable are unpredictable and the
experiments become random.

We have already learnt about random experiments and formation of sample

spaces. In a random experiment, we are more interested in, x number associated with
the outcomes in the sample space rather than the individual outcomes. These numbers
vary with different outcomes of the experiment. Hence it is a variable. That is, this value
is associated with the outcome of the random experiment. To deal with such situation we
need a special type of variable called random variable.

9.1 Definition of random variable

Definition X���

Let S be the sample space of a random

experiment. A rule that assigns a single real �
number to each outcome (sample point) of the –3
x +3
Real line
random experiment is called random variable. Sample space Fig.9.1

In other words, a random variable is a real valued function defined on a sample

space S that is with each outcome ω of a random experiment there corresponds a unique
real value x known as a value of the random variable X. That is X(ω ) = x.

Generally random variables are denoted by upper case alphabets like X, Y, Z …

and their values or realizations are denoted by the corresponding lower case letters. For
example, if X is a random variable, the realizations are x1, x2 …

Example 9.1

Consider the random experiment of rolling a die.

The sample space of the experiment is S={1, 2, 3, 4, 5, 6}

Let X denotes the face of the die appears on top.

1 2 3 4 5 6
Fig.9.2
 The assigning rule is

X(1) = 1, X(2) = 2, X(3) = 3, X(4)=4, X(5)=5 and X(6)=6

Hence the values taken by the random variable X are 1,2,3,4,5,6. These values are
also called the realization of the random variable X.

Example 9.2

Random experiment : Two coins are tossed simultaneously.

Sample space : S={HH, HT, TH, TT}

Assigning rule : Let X be a random variable defined as the number of heads

comes up.

Sample Point ω HH HT TH TT

X (ω ) 2 1 1 0

Here, the random variable X takes the values 0, 1, 2 .

Example 9.3

Experiment : Two dice are rolled simultaneously.

Sample space : {(1, 1),(1, 2),(1, 3),…(6, 6)}

Assigning rule : Let X denote the sum of the numbers on the faces of dice

then Xij = i + j, Here, i denotes face number on the first die and j
denotes the face number on the second die.

Then X is a random variable which takes the values 2, 3, 4 . .… 12.

That is the range of X is {2, 3, 4…… 12}

9.2 Discrete and Continuous random variables

Random variables are generally classified into two types, based on the values they
take such as Discrete random variable and Continuous random variable.

9.2.1 Discrete random variable

A random variable is said to be discrete if it takes only a finite or countable infinite
number of values.
Example 9.4

Consider the experiment of tossing a coin

ΝΟΤΕ
If X (Head) =1, X (Tail) = 0
Example 9.1, 9.2, and 9.3 are
Then X takes the values either 0 or 1 the random variables taking
finite number of values,
This is a discrete random variable.
therefore they are discrete.

Example 9.5

Consider the experiment of tossing a coin till head appears.

Let random variable X denote the number of trials needed to get a head. The values
taken by it will be 1, 2, 3, ..

It is discrete random variable taking countable infinite values.

9.2.2 Continuous random variable:

A random variable X is said to be continuous, if it takes values in an interval or
union of disjoint intervals. (A rigorous definition is beyond the scope of the book).

Example 9.6

If X is defined as the height of students in a school ranging between 120 cms and
180 cms, Then the random variable X is {x/120 cms < x < 180 cms } is a continuous
random variable.

Example 9.7

Let the maximum life of electric bulbs is 1500 hrs. Life time of the electric bulb is the
continuous random variables and it is written as X = {x/0 ≤ x ≤ 1500}

9.3 Probability mass function and probability density function

P(x)
A probability function is associated with each value 3
of the random variable. This function is used to compute 8

probabilities for events associated with the random variables. 2

8
The probability function defined for a discrete random
1
variable is called probability mass function. The probability 8
function associated with continuous random variable is called
0 1 2 3 x
probability density function.
Fig.9.3
9.3.1 Probability Mass Function.
If, X is a discrete random variable taking values x1, x2 …. xn with respective
probabilities p(x1), p(x2) ….. p(xn) such that

(i) p(xi) ≥ 0, 6 i ( non- negative ) and (ii) /i = 1 p ^ xih = 1 then p(x) is known as the
n

probability mass function (p.m.f) of the discrete random variable X.

The pair {xi, p(xi); i = 1, 2, 3, ... } is known as prabability distribution of X.

Example 9.8

A coin is tossed two times. If X is the number of heads, find the probability mass
function of X.

Solution:

Since the coin is tossed two times, the sample space is S={HH, HT, TH, TT}

If X denotes the numbers of heads, the possible values of X are 0,1,2 with the
following
1
P(X = 0) = P (getting no head) = 4 NOTE
2 1
P(X = 1) = P (getting one head) = 4 = 2 Probabilities are non
1 negative and the
P(X = 2) = P (getting two heads) = 4
total is 1.
The probability distribution of X is

X 0 1 2
1 1 1
p(X = x) 4 2 4

Example 9.9

In example 9.3 the probability mass function of X is given in the following table

X 2 3 4 5 6 7 8 9 10 11 12
1 2 3 4 5 6 5 4 3 2 1
P(x) 36 36 36 36 36 36 36 36 36 36 36
The above table may be called as the probability distribution function of X.

9.3.2 Probability Density Function

When the random variable is continuous in the co-domain it is spread over it. A
function f (x) is defined on real line and satisfying the following conditions :
 (i) f ]xg $ 0 , 6x (ii) #- 33 f]xgdx = 1 is called the probability density function
(p.d.f) of X.

Remark:

(i) Every integrable function satisfying the above

two conditions is a probability density function
of some random variable X

(ii) The probability that the value of X lies between a b

two points ‘a’ and ‘b’ is P ]a < X < bg = # f ]x g dx

b Fig.9.4
a

(iii) If X is discrete random variable then for any real x, P(X = x) need not be
zero.However in the case of continuous random variable P(X = x) = 0 holds
always. P ] X - ag = # f ]xg dx = 0
a
a

(iv) If X is a continuous random variable then for any a < b

P(a < X < b) = p(a ≤ X < b) = p(a < X ≤ b) = p(a ≤ X ≤ b)

Example 9.10

A continuous random variable X has probability density function given by

Ax3, 0 < x < 1
f ]xg = * . Find A.
0 , otherwise

Solution:

Since f(x) is a p.d.f NOTE

#- 33 f]xgdx = 1 Total probability of

continuous random
#0 1 Ax3 dx = 1 variable is 1 within the
A certain interval.
& 4 = 1
& A = 4

Example 9.11

Verify whether the following function is a probability density function

2x
f ]xg = * 9
, 0<x<3
0 , elsewhere
Solution :

#- 33 f]xgdx = #0 3 29x dx
 9 :2D = 1
2 9
&

It is to be noted that (i) f ]xg $ 0, 6x (ii) #- 33 f]xgdx =1

Hence, f(x) is a p.d.f.

9.4 Distribution function and its properties P(x)

We get the probability of a given event at a

particular point. If we want to have the probability upto
the point we get the probability P(X ≤ x). This type of
probability is known as probability mass function. We
can also find how the probability is distributed within
certain limits. [P (X < x) or P (X > x) or P (a < x < b)]. 0 1 2 3 4 5 6 7 x

Fig.9.5

9.4.1 Distribution Function for discrete random variable

Definition: Let X be a random variable ,the cumulative distribution function (c.d.f)
of a random variable X is defined as F(x) = P(X≤x), 6 x. It is called simply as distribution
function.

9.4.2 Properties: NOTE

(i) 0≤F(x)≤1, 6 x , (non –negative) In the case of discrete random

variable X
(ii) F(x)≤F(y), 6 x<y , (non –decreasing) x
(i) F(x) = / P ]x = r g
(iii) lim F ]x + hg = F(x), 6 x , (F(x) is right r =- 3
h"0 (ii) P(X = x) = F(x) – F (x–1)
continuous)

(iv) F ]- 3g = lim F ]x g = 0
x "-3
(v) F ]3g = lim F ]xg = 1
x"3

9.4.3 Distribution Function for continuous random variable

(i) F(x) = #- 3x f]xgdx That is pdf can be obtained from

(ii) f(x) = F l]xg

distribution function by differentiating
the distribution function and the
(iii) P ]a < X < bg = P ]a # X < bg distribution function can be obtained

= P ]a < X # bg = P ]a # X # bg
from pdf by integrating the pdf over
the given range

f ]xg dx
b
= #a
Properties
(i) F(x) is a non decreasing function of x

(ii) 0<F(x)<1 -3 < x < 3

(iii) F ]- 3g = 0,

(iv) F ]3g = 1

(v) For any real constant a and b such that a < b, p( a < X ≤ b) = F(b) – F(a)

(vi) f(x) = dx ^F ]xgh i.e., f(x) = F l]xg

d

Example 9.12

A random variable X has the following probability mass function

X 0 1 2 3 4 5 6
P(X = x) a 3a 5a 7a 9a 11a 13a

(i) Find the value of ‘a ’

(ii) Find the c.d.f F(x) of X

(iii) Evaluate : (a) P(X $ 4) (b) P(X < 5) (c) P(3 ≤ X ≤6)

(iv) P(X = 5) using F(x)

Solution:

(i) Since P(X = x) is probability mass function / P ] X = xg = 1

i.e., P(X = 0) + P(X =1) + P(X = 2) + P(X = 3) + P(X = 4) + P(X = 5) + P(X = 6) = 1

a + 3a + 5a + 7a + 9a + 11a + 13a = 1
1
49a = 1 => a = 49

(ii) Hence the c.d.f is

X 0 1 2 3 4 5 6
1 3 5 7 9 11 13
P(x) 49 49 49 49 49 49 49
1 4 9 16 25 36 49
F(x) 49 49 49 49 49 49 49 = 1
 (iii) (a) P(X ≥ 4)

= P(X = 4) + P(X = 5) + P(X = 6)

= 9a + 11a + 13a

= 33a
1 33
= 33 # 49 = 49

(b) P(X<5) = 1– P (X ≥ 5) = 1– [P (X = 5) + P (X = 6) ]

= 1 – [11a + 13a]

= 1 – 24a
24
= 1 – 29
25
= 29

(c) P(3 ≤ × ≤6) = P (X = 3 ) + P (X = 4 )+ P (X = 5 ) + P (X = 6)

= 7a + 9a + 11a + 13a

= 40
40
a = 49

(iv) P(X =5) = F(5) – F(5 – 1)

36 25
= 49 - 49
11
= 49 .

Example 9.13

Let X be a random variable with p.d.f

x
f ]xg = * 2
; 0<x<2
0 ; otherwise
(i) Find the c.d.f of X , (ii) Compute P b 2 < X # 1 l ,
1
(iii) Compute P(X=1.5)

Solution:

(i)The c.d.f of X : F(x) = P(X ≤ x)

x
= # f ]xg dx
-3
 x
# x x2
= 2 dx = 4
0
Z] 0 if x < 0
]]
] x2
Hence, F(x) = []
]] 4 if 0 < x < 2
]1 if x $ 2
\ NOTE
P b 2 < X # 1 l = F ]1 g - F b 2 l
1 1
(ii)
For a continuous
1 1 3 random variable
= 4 - 16 = 16
probability at a
This probability can be computed using p.d.f particular point is 0

= 2 b x2 l1
#1 1 2x dx 1 2 1
2 2
3
= 16

(iii) P (X = 1.5) = 0

9.5 Joint and marginal probability mass functions

In real life situations we may be observed two or more random variables on the
individuals simultaneously. For instance, blood pressure and cholesterol for each individual
are measured simultaneously. In such cases we require the concept of bi-variate random
variable represented by ( X, Y ), where X and Y are univariate random variables.

9.5.1 Definition (Joint p.m.f)

Let (X, Y) be a discrete bivariate random variable. Then p (x, y) is called the joint
probability mass function of (X, Y) if the following conditions are satisfied.

p (x, y) ≥ 0 6 x, y NOTE
/ x, y p^x, yh = 1 p (x, y) = p (X = x, Y =y)

Definition (Marginal Probability Mass Function)

Given a joint probability mass function p (x, y), then p (x) = / y p^x, yh is called
marginal probability mass function of X. Similarly p (y) = / x p^x, yh is called the marginal
probability mass function of Y.

Example 9.14

There are 10 tickets in a bag which are numbered 1, 2, 3, ...10. Two tickets are
drawn at random one after the other with replacement.
 Here, random variable X denotes the number on the first ticket and random variable
Y denotes the number on the second ticket.

9.5.2 Joint and marginal probability density functions

As we defined in section 9.5.1 the joint probability mass function, we define the
joint probability density function.

Definition:
Let (X, Y) be a bivariate continuous random variables. The function f (x, y) is
called a bivariate probability density if the following conditions are satisfied.

(i) f (x, y) ≥ 0 6 x, y
3 3
(ii) # # f ^ x, y h dx dy = 1
-3 -3

The marginal probability density function of X is given by

3
g(x) = # f ^ x, y h dy
-3
and the marginal probability density function of Y is given by
3
h(y) = # f ^ x, y h dx
-3

Example 9.15

1 ^x + y h
Prove that the bivariate function given by f(x, y) = * 8
, 0 < x, y # 2
0 , otherwise
Proof:
3 3
If f is a probability density function # # f ^ x, y h dx dy = 1
2 2 -3 -3
1
= 8 # # ^x + yhdx dy
0 0
2 2 2 2
= 8f ydx dy p
1 # # xdx dy + # #
0 0 0 0
2 2 2 2

= 8 * # > # xdx Hdy + # ># ydy Hdx 4

1
0 0 0 0
2 2
= 8 > # 2dy + # 2dxH
1
0 0
 = 8 8^2y h0 + ]2xg02B
1 2

= 8 6]4 - 0g + ]4 - 0g@
1

1
= 8 #8 = 1

Therefore, f (x, y) is a probability density function.

Example 9.16

x 2 y , 0 # x # 1, 0 # y # 2
Joint p.d.f. of X,Y is f(x, y) = * then find the marginal
0 , elsewhere
density function of X and Y.

Solution:
3
f (x) = # f ^ x, y h dy
-3
2
f (x) = # x2 ydy
1
2
y2
= x2 < F = x2 : 2 - 2 D
4 1
2 1
3 3
= x2 # 2 = 2 x2
3
f (y) = # f ^ x, y h dx
-3
1
f (y) = # x2 ydy
0
3 1 y
= y c x3 m = y b 3 l = 3
1
0

Marginal density function of X

3 2
f(x) = * 2
x , 0 # x # 1,
0 , elsewhere
Marginal density function of Y

y
f(x, y) = * 3 , 0 # y # 2
0 , elsewhere

Example 9.17

4 x - 2 y , 0 # x # 3, 0 # y # 2
Joint p.d.f. of X,Y is f(x, y) = ) Find the marginal
0 , elsewhere
density function of X and Y
Solution:
3
f (x) = # f ^ x, y h dy
-3
2
f (x) = # ^4x - 2yhdy
1
2
2y 2
= <4xy - F
2 1

= 64x ]2 - 1g - ]4 - 1g@ = 4x - 3
3
f (y) = # f ^ x, y h dx
-3
3
f (y) = # ^4x - 2yhdx
1

= b 42x - 2xy l
2 3

= _2x2 - 2xy i1
3

= 72 ]9 - 1g - 2y ]3 - 1gA

= 2 # 8 - 6y + 2y

= 16 - 4y

Marginal density function of X

4x - 3 , 1 # x # 3
f(x) = )
0 , elsewhere

Marginal density function of Y

16 - 4y , 1 # y # 2
f(x, y) = )
0 , elsewhere

9.6 Mathematical expectation

Probability distribution gives us an idea about the likely value of a random variable
and the probability of the various events related to random variable. Even though it is
necessary for us to explain probabilities using central tendencies, dispersion, symmetry
and kurtosis. These are called descriptive measures and summary measures. Like frequency
distribution we have to see the properties of probability distribution. This section focuses
on how to calculate these summary measures. These measures can be calculated using
(i) Mathematical Expectation and variance.
(ii) Moment Generating Function .
(iii) Characteristic Function.

9.6.1 Expectation of Discrete random variable

Let X be a discrete random variable which takes the values x1, x2 ….. xn with
respective probabilities p1, p2 ….. pn then mathematical expectation of X denoted by E(X)
is defined by

E(X) = x1 p1 + x2 p2 +… + xn pn

= /ni = 1 xi pi where / pi = 1
Sometimes E(X) is known as the mean of the random vairable X.

Result:
If g(X) is a function of the random variable X, then E g(X) = / g ]xg p ] X = xg
Properties:
(i) E(c) = c where c is a constant

Proof :
E(X) = / xi pi
E(c) = / cpi = c / pi = c # 1 = c
(ii) E ]cX g = cE(X), where c is a constant

Proof:
E(X) = / xi pi
E(X) = / cxi pi
= c / xi pi

= c E(X).

(iii) E(aX+b) = aE(X)+b

Variance of Discrete random variable
Definition: In a probability distribution Variance is the average of sum of squares of
deviations from the mean. The variance of the random variable X can be defined as.

Var ( X ) = E (X – E(X))2

= E (X2 ) – ( E(X))2

Example 9.18

When a die is thrown X denotes the number turns up. Find E(X), E(X2) and Var(X).

Solution:

Let X denoted that number turns up in a die.

1
X takes the values 1, 2, 3, 4, 5, 6 with probabilities 6 for each.

Therefore the probability distribution is

X 1 2 3 4 5 6

P(x) 1 1 1 1 1 1
6 6 6 6 6 6
E (X) = ∑xi pi

E (X) = x1 p1 + x2 p2 + … + x6 p6
1 1 1
E (X) = 1 × 6 + 2 × 6 + … + 6 × 6
1
= 6 (1+2+3+4+5+6)
7
= 2

E (X2) = ∑xi 2pi

E (X2) = x1 2p1 + x22 p2 + … + x6 2p6

E (X2) = b12 6 l + b 22 6 l + ... b 62 6 l

# 1 # 1 # 1

1
= 6 (1+4+9+16+25+36)
1 91
= 6 (91) = 6

Var (X) = E (X2 ) – ( E(X))2

= 6 -b 2 l
91 7 2
 91 49
= 6 - 4
35
Var(X) = 12

Example 9.19

The mean and standard deviation of a random variable X are 5 and 4 respectively
Find E (X2)

Solution:

Given E (X) = 5 and v = 4

` Var (X) = 16

But, Var (X) = E(X2) – [E(X)]2

16 = E (X2) – (5)2

E(X2) = 25 + 16 = 41

Example 9.20

A player tosses two coins, if two head appears he wins ` 4, if one head appears he
wins ` 2, but if two tails appears he loses ` 3. Find the expected sum of money he wins?

Solution:

Let X be the random variable denoted the amount he wins.

The possible values of X are 4, 2 and –3

1
P(X = 4) = P(getting two heads) = 4
1
P(X = 2) = P(getting one head) = 2
1
P(X = 3) = P(getting No heads) = 4

Probability distribution is

X 4 2 -3

P(X=x) 1 1 1
4 2 4
E (X) = ∑xi pi

E (X) = x1 p1 + x2 p2 + x3 p3
1 1 1
E (X) = 4 × 4 + 2 × 2 –3 × 4 = 1.25
Example 9.21

Let X be a discrete random variable with the following probability distribution

X –3 6 9

P(X=x) 1 1 1
6 2 3
Find mean and variance
Solution:

Mean = E(X) = ∑xi pi

= x1 p1 + x2 p2 + x3 p3
1 1 1
= -3× 6 + 6 × 2 + 9 × 3
11
= 2

E (X2) = ∑xi 2pi

= x1 2p1 + x22 p2 +x3 2p3

= b - 32 6 l + b 62 2 l + ... b 92 3 l
# 1 # 1 #1

3
= 2 +18+27
93
= 2

Van (X) = E (X2 ) – ( E(X))2

= 2 -b 2 l
93 11 2

93 121
= 2 - 4
186 - 121
= 4
65
= 4
11
Mean = 2 ,
65
Var (X) = 4

9.6.2 Expectation of a continuous random variable

Let X be a continuous random variable with probability density function f (x) then
the mathematical expectation of X is defined as
 3
E (X) = # xf (x) dx provided the integral exists
-3
3
E [ g(X)] = # g (x) f (x) dx
-3 NOTE
Results:
If g(X) is a function of a
(1) E(c) = c where c is constant random variable and E
[g(X)] exists then
Proof :
3
By definition, E(X) = # xf (x) dx
-3
3 3
E(c ) = # cf (x) dx = c # f (x) dx
-3 -3
3
= c×1=c fa # f (x) dx = 1 p
-3
(2) E (aX) = a E(X)

Proof :
3 3
E (aX) = # axf (x) dx = a # xf (x) dx
-3 -3

= a E (X)

Example 9.22

The p.d.f. of a continuous random variable X is given by

x
f (x) = * 2
, 0<x<2
find its mean and variance
0, elsewhere
Solution:
3
E (X) = # xf (x) dx
-3
2
x b 2 l dx
x
E(X) = #
0
2
1 #
= 2 x2 dx
0

1 8 3 B2
= 2 x 30
 1 8
= 2 [3–0]
4
= 3
3
E(X2) = # x2 f (x) dx
-3
3 NOTE
2b
2 l dx
x
= # x
-3 The following results are true in both
discrete and continuous cases.
1 8x 4 4B0
2
= 2 (i) E ( 1/X) ≠ 1/ E(X)
1
= 8 [16-0] (ii) E [log (X)] ≠ log E (X)
= 2 (iii) (E (X2) ≠ [E(X)]2

Variance (X) = E(X2) – [E(X)]2

4
= 2–( 3 )2
16 2
= 2– 9 =9

So far we have studied how to find mean and variance in the case of single random
variables taken at a time but in some cases we need to calculate the expectation for a linear
combination of random variables like aX + bY or the product of the random variables
cX × dY or involving more number of random variables. So here we see theorems useful
in such situations.

9.6.3 Independent random variables

Random variables X and Y are said to be independent if the joint probability density
function of X and Y can be written as the product of marginal densities of X and Y.

That is f (x,y)= g(x) h(y)

Here g(x) marginal p.d.f. of X

h(y) marginal p.d.f. of Y

9.7 Addition and Multiplication Theorem on Expectations

9.7.1 Addition Theorem on Expectations

1. Statement for Discrete random variable

If X and Y are two discrete random variables then
E (X+Y) = E(X) + E(Y)
Proof

Let the random variable X assumes the values x1, x2 ... xn with corresponding
probabilities p1, p2…..pn, and the random variable y assume the values y1, y2 ……. y m with
corresponding probabilities p1, p2…..pm

By definition,

E (X) = / xipi where / pi = 1

E(Y) = / y jpj / pj = 1

Now, E (X+Y) = /in= 1 /mj = 1 pij (xi + y j)

= /in= 1 /mj = 1 pij xi + /nj = 1 pij (xi + y j)

= /in= 1 /mj = 1 pij xi + /in= 1 /im= 1 pij y j

= /in= 1 xi pi + /mj = 1 y j p j

= E (X) + E (Y)

2. Statement for Continuous random variable

Let X and Y are two continuous random variables with probability density functions
f(x) and f(y) respectively. Then E (X + Y) = E(X) + E(Y)

Proof:
We know that
3
E(X) = # xf (x) dx and
-3
3
E(Y) = # yf (y) dy
-3
3 3
E(X+Y) = # # ^x + y h f ^x, y h dxdy
-3 -3
3 3
xf f ^ x, y h dy p dx + yf f ^ x, y h dx p dy
3 3
= # # # #
-3 -3 -3 -3

3 3
= # xf (x) dx + # yf (y) dy
-3 -3

= E (X) + E (Y)
 E (aX + b) = a E (X) ! b
3
E (aX±b) = # (ax ! b) f (x) dx
-3
3 3
=a # xf (x) dx ! b # f (x) dx
-3 -3
3
= a E (X) ! b # 1 as # f (x) dx = 1
-3

= a E ((X) ± b)

Remarks:
1. Statement: E (aX+b) = aE(X)+b where a and b are constants.

Proof: E(aX+b) = E(aX)+E(b) by property 3

= aE(X)+b by property 2

Similarly E(aX–b) = aE(X)–b

6aE ]xg + b@
2. Statement: E b c l =
ax + b
c
]aX + bg E ]aX + bg
Proof: E c = c
[aE (x) + b]
= c

3. Statement: E(X-X ) =0

Proof: E ( X - X ) = E(X) - E( X ) since E(X)= X

= X - X =0 X is a constant

Example 9.23

Find the expectation of the sum of the number obtained on throwing two dice.

Solution:

Let X&Y denote the number obtained on the I and II die respectively. Then each
of them is a random variable which takes the value 1,2,3,4,5 and 6 with equal probability
1
6
 E(X) = /xi pi
1 1 1
= 1× 6 +2× 6 +.....+6× 6

1 + 2 + 3 + 4 + 5 + 6 21 7
= 6 = 6 =2

Similarly,
7
E(Y) = 2

Thus the expectation of the numbers obtained on two dices.

X+Y takes the values 2, 3…12 with their probability given by

x 2 3 4 5 6 7 8 9 10 11 12

P(x) 1 2 3 4 5 6 5 4 3 2 1
36 36 36 36 36 36 36 36 36 36 36
1 2 1
E(X+Y) = 2 # 36 + 3 # 36 + ... + 12 # 36

2 + 6 + 12 + 20 + 30 + 42 + 40 + 36 + 30 + 22 + 12
= 36 =7
7 7
E(X)+E(Y) = 2 + 2 = 7

` E(X+Y) = E(X) + E(Y)

Example 9.24

Let X and Y are two random variables with p.d.f given by

4xy, 0 # x # 1
f(x,y) = )
0 otherwise
and their marginal density functions as
2x, 0 # x # 1
f(x)= )
0 otherwise

2y, 0 # x # 1
and g(y)= )
0 otherwise
prove that E(X+Y)= E(X)+E(Y)

Solution:
3 3
E(X+Y) = # # ^x + y h f ^x, y h dxdy
-3 -3
1 1
= # # ^x + yh4xy dxdy
0 0
 1 1 1 1
= 4> # #x # # xy2 dx dyH
2
y dx dy +
0 0 0 0
1 1 1 1
= 4> # f #x dx p ydy + #f# y2 dy p xdxH
2

0 0 0 0
1 1
# 1 # 1
= 4[ 3 y dy + 3 x dx ]
0 0
1 1
= 3 > # y dy + # x dxH
4
0 0
4 1 1 4
= 3 [2 + 2] = 3 ... (1)
3
E(X) = # xf (x) dx
-3
1
= # x # 2x dx
0
1
=2 # x2 dx
0

E(X) = 2 : 3 D = 3
1 2

3
E(Y) = # yf (y) dy
-3
1
= # y # 2y dy
0
1
=2 # y2 dy
0

E (Y) = 2 : 3 D = 3
1 2

2 2 4
E(X)+E(Y) = 3 + 3 = 3 ... (2)

From 1&2

E(X+Y) = E(X)+E(Y)

9.7.2 Multiplication Theorem on Expectation

Discrete random variable:

Statement: If X and Y are two independent variables then E(XY) = E (X) E(Y)
 Proof:

E (XY) = /in= 1 /mj = 1 xi y j pij If X and Y are independent

n m
= /i = 1 / j = 1 xi y j pi p j Pij = pi p j

= a /i = 1 xi pi k b / j = 1 yi pi l
n m

= E(X) E(Y)

Continuous random variable:

Statement: If X and Y are independent random variables Then E (XY) = E(X) E(Y)

Proof:
3 3
Now E(XY) = # # xy f ^ x, y h dxdy
-3 -3
3 3
== # xf (x) dx # yf (y) dy
-3 -3

= E (X) E (Y)

Example 9.25

Two coins are tossed one by one. First throw is considered as X and second throw
is considered as Y following joint probability distribution is given by,

X 1 0 Total
Y
0.25 0.25 0.5
1

0.25 0.25 0.5

0

0.5 0.5 1
Total

[getting Head is taken as 1 and Tail is taken as 0]

Verify E(XY)= E(X) E(Y)

Solution:

A random variable XY can take the values 0 and 1

 E (XY) = Σ Σ xy p(x,y)

=1 × 0.25 + 0 × 0.25 + 0×0.25 +0×0.25

=0.25

E (X) = Σ x pi

= 1 ×0.5+0×0.5

= 0.5

E (Y) = Σ y pj

= 1 ×0.5+0×0.5 = 0.5

E(X)×E(Y) = 0.5 ×0.5 = 0.25

E (XY) = E (X) E(Y)

[It is applicable only when X and Y are independent]

Example 9.26

The independent random variables X and Y have the p.d.f given by

4ax, 0 # x # 1
f (x) = )
0, otherwise

4by, 0 # y # 1
f (y) = )
0, otherwise
Prove that E(XY) = E(X) E(Y)

Solution:

X and Y are independent

f(x,y) = f(x) × f(y)

f(x,y) = 4ax×4by
16abxy, 0 # x, y # 1
f(x,y) = )
0, otherwise
3 3
E(XY) = # # xy f ^ x, y h dxdy
-3 -3
1 1
= # # xy # 16abxydxdy
0 0
 1 1
= 16ab # > # x2 dxHy2 dy
0 0
1
# 1 2
= 16ab 3 y dy
0
1 1 16ab
= 16ab× 3 # 3 = 9 →1
3
E(X) = # xf (x) dx
-3
1 1
= # x # 4ax dx = 4a # x2 dx
0 0
1 4a
= 4a ( 3 ) = 3
3
E(Y) = # yf (y) dy
-3
1
= # y # 4by dy
0
1
=4a # y2 dy
0
1 4
=4b ( 3 ) = b

4a 4b
E(X) E(Y) = 3 # 3

16ab
= 9 →2
From 1 and 2,

E(XY) = E(X) E(Y)

9.8 Moments
Another approach helpful to find the summary measures for probability distribution
is based on the ‘moments’.We will discuss two types of moments.

(i) Moments about the origin. (Origin may be zero or any other constant say A ).
It is also called as raw moments.

(ii) Moments about the mean is called as central moments.

9.8.1.Moments about the origin
Definition:
If X is a random variable and for each positive integer r (r = 1, 2, ...) the rth raw
moment about the origin ‘0’ is n'r = E(Xr) = / xi r pi

First order raw Moment

Put r = 1, in the definition

n'1 = E(X) = / xi pi = X

This is called the mean of the random variable X. Hence the first order raw moment
is mean.

Second order raw moment

Put r = 2 then

n'2 = E(X2)= / xi2pi

This is called second order raw moment about the origin.

9.8.2. Moments about the mean (Central moments)

Definition:
For each positive integer r, (r = 1, 2, ...), the rth order central moment of the random
variable X is nr = E ( X - X )r

First order central moment:

put r = 1 in the definition, then

n1 = E ( X - X ) = 0 (always)

Remark:
The algebraic sum of the deviations about the arithmetic mean is always 0

Second order central moment;

Put r = 2 in definition
n 2 = E ( X - X )2

= E ( X2 – 2 X # X + X 2 )
 = E ( X2) – 2E (X) E ( X ) +E ( X )2

= E(X2) – 2 X X + ( X )2

= E(X2) – ( X )2

= E(X2) – [E(X)]2 where X is constant and E(X) = X

n2 = n'2 - (n'1) 2

This is called the 2nd central moment about the mean and is known as the variance
of the random variable X.

i.e., Variance = Var (X) = n2 = n'2 - (n'1) 2

Standard Deviation (S.D) = v = variance

Some results based on variance:

(i) Var (c) = 0 i.e. Variance of a constant is zero

(ii) If c is constant then Var (cX)=c2 Var(X)

(iii) If X is a random variable and c is a constant then Var (X ± c) = Var (X)

(iv) a and b are constants then Var (aX ± b) = a2 Var (X)

(v) a and b are constants then Var (a ± bX) = b2 Var (X).

(vi) If X and Y are independent random variables then Var (X + Y)

= Var (X) + Var(Y)

9.8.3. Moment generating function(M.G.F.)

Definition:

A moment generating function (m.g.f) of a random variable X about the origin is

denoted by M x(t) and is given by

M x(t) = E (etx) , |t| <1

(i) Mx(t) = / (etx) p (x) for a discrete distribution

3
(ii) Mx(t) = # etx f ]xg dx for a continuous distribution
-3

For a random variable X to find the moment about origin we use moment generating
function.
 E (Xr) = M xr (t) at t=0 M xr (t)

rth differential of Mx(t) is

n2 = n2l - (n1l ) 2

To find the rth moment of X about the origin.

We know that
Mx(t) = E (etX)
tX (tX)
2
(tX) 3 (tX) r
= E [1+ 1! + 2! + 3! + ... + r! + ...]

E (tX) E (tX) 2 E (tX) 3 E (tX) r

= 1+ 1! + 2! + 3! + ... + r! + ...
E (X) t2 E (X2) tr E (Xr)
= 1+t 1! + 2! + ... + r! + ...
n' n'
= 1 + t n'1 + t2 2!2 + ... + tr r!r + ...
n'
Mx (t) = /r3= 0 tr r!r

tr
From the series on the right hand side, nr ' is the coefficient of r! in Mx (t) .

Since Mx (t) generates moments of the distribution and hence it is known as moment
generating function.

Using the function, we can find mean and variance by using the first two raw
moments.
n2 = n'2 - (n'1) 2

9.8.4. Characteristic function

For some distribution, the M.G.F does not exist. In such cases we can use the
characteristic function and it is more servicable function than the M.G.F.

Definition:
The characteristic function of a random variable X, denoted by z x (t) , where
z x (t) = E (ei t X) then
3
z x (t) = E (e i tX
)= # eitx f ]xg dx , for continuous random variable
-3
z x (t) = /eitx p (x) , for discrete random variable

where i = -1