Subject Name with Code-

MODULE-V
Complex Analysis, Probability
and Statistical Methods Joint Probability Distribution and Sampling
18MAT41
Theory

Mrs. Divya K.
Assistant Professor,
Dept. of Mathematics,
ATMECE, Mysuru.

<Dept. Name> <Faculty Name>

 INTRODUCTION
We have discussed probability distribution associated with a single random variable. The same can be
generalized for two or more random variables. We discuss probability distributions associated with two
random variables referred to as a joint distribution.
JOINT DISTRIBUTION AND JOINT PROBABILITY DISTRIBUTION
  If X & Y are two discrete random variables, we define the joint probability function of X & Y by
P (X = x, Y = y) = f (x,y)
Where f(x,y) satisfy conditions, f(x,y) ≥0 and ∑x ∑y f(x,y) = 1
The second condition means that the sum over all the values of x and y is equal to one.
Suppose X = and Y = {then P ( X = xi , Y = yi) denoted by J ij.

It should be observed that f is a function on the Cartesian product of the sets X and Y as we have
X×Y=

<Dept. Name> <Faculty Name>

f is also referred to as joint probability density function of X and Y in the respective order. The set of values of this
function f(xi, yi) = J ij for i = 1,2,...m, j= 1,2,...n is called the joint probability distribution of X and Y. These values
are presented in the form of a two way table called the joint probability table.
X Y ... Sum
  f(x1)
... f(x2)
... ...   ... ... ...
... f(xm)
g(y1) g(y2)   g(yn) 1

<Dept. Name> <Faculty Name>

MARGINAL
  PROBABILITY DISTRIBUTION
 
In the joint probability table are the sum of horizontal entries and are the sum of vertical entries in the joint
probability distribution table. These are called marginal probability distribution of X and Y respectively.
 
INDEPENDENT RANDOM VARIABLES

The discrete random variable X and Y are said to be independent random variables if
P ( X = xi , Y = yj) = P ( X = xi ) . P ( Y= yj )
i.e f(xi) g(yj) = Jij

Expectation,
  Variance, Covariance and Correlation
Expectation
= E(X) =
= E(Y) =
= E(XY) = 

<Dept. Name> <Faculty Name>

 Variance
= E ( X 2)
= E ( Y2)
 
Covariance
COV (X,Y) = E (XY) – E(X) E(Y)
 
Correlation
Correlation of X and Y = =
NOTE:
If X and Y are independents, E(X , Y) = E(X) E(Y) and hence COV (X , Y) = 0 =

<Dept. Name> <Faculty Name>

PROBLEMS
1. The joint probability distribution of two random variables X and Y is as follows.

X Y -4 2 7

1 1/8 1/4 1/8

5 1/4 1/8 1/8

  Compute the following

(a) E(X) and E(Y) (b) E( XY) (c) and (d) COV (X , Y) (e)

Solu: The distribution is obtained adding the all the respective row entries
and also the respective column entries.
Distribution of X : Distribution of Y :

1 5 -4 2 7
1/2 1/2 3/8
3/8 3/8
3/8 1/4
1/4

<Dept. Name> <Faculty Name>

 a) E(X) = = (1) (1/2) + 5 (1/2) = 3 =

E(Y) = = (-4) (3/8) + 2(3/8) + 7(1/4) = 1 =

 
 b) E (XY) = = (1) (-4) (1/8) + (1) (2) (1/4) + (1) (7) (1/8) + (5) (-4) (1/4) + (5) (2) (1/8) + (5) (7) (1/8)

= 3/2

  c) and

 Now E (X2) = = (1) (1/2) + 25(1/2) = 13

 
E (Y2) == (16) (3/8)+ (4)(3/8) + (48)(1/4) = 79/4

  Hence = 13 – (3)2 = 4 and = (79/4) – (1)2= 75/4

<Dept. Name> <Faculty Name>

Thus
  = 2 and = = 4.33
 
d) COV (X,Y) = E (XY) – E(X) E(Y)
= (3/2) – 3 (1) = - 3/2
 
e) = = = - 0. 1732.

<Dept. Name> <Faculty Name>

2. The joint probability distribution table for two random variables X and Y is as follows.

X Y -2 -1 4 5

1 0.1 0.2 0 0.3

2 0.2 0.1 0.`1 0

 Determine the marginal probability distributions of X and Y. Also compute

(a) Expectations of X , Y and XY
(b) S.D’s of X,Y
(c) covariance of X and Y
(d) Correlation of X and Y
Further verify that X and Y are dependent random variables and also find
 
Solution: Marginal distributions of X and Y are got by adding all the respective row entries and the respective column
entries.

<Dept. Name> <Faculty Name>

 xi 1 2 yj -2 -1 4 5
f(xi) 0.6 0.4 g(yj) 0.3 0.3 0.1 0.3

  (a) E(X) = = (1)(0.6) + (2)(0.4) = 1.4 =

  E(Y) = = (-2) (0.3) + (-1)(0.3) + 4 (0.1) + 5 (0.3) = 1=

 E (XY) = = (1)(-2)(0.1) + (1) (-1) (0.2) + (1) (4) (0) + (1) (5) (0.3) + (2) (-2) (0.2) + (2) (-1) ( 0.1)
  + (2) (4) (0.1) + (2)(5) (0)

= 0.9

  b) and

<Dept. Name> <Faculty Name>

 Now E ( X2) = = (1)(0.6) + (4) (0.4) = 2.2
 
E ( Y2) = = (4)(0.3) + 1(0.3) + 16(0.1) + 25(0.3) = 10.6
 

 Hence = 2.2 – (1.4)2 = 0.245 and = (10.6) – (1)2= 9.6

 
Thus = 0.49 and =

 c) COV (X,Y) = E (XY) – E(X) E(Y)

= 0.9 – 1.4(1) = - 0.5

  d) = = = - 0.3.

<Dept. Name> <Faculty Name>

 If X and Yare independent random variables we must have
f(xi) g(yj) = Jij
It can be seen that f(x1)g(y1) = (0.6)(0.3) = 0.18 and J11 = 0.1
i.e f(x1)g(y1) J11
Hence we conclude that X and Y are dependent random variables.
   
 

<Dept. Name> <Faculty Name>

  

<Dept. Name> <Faculty Name>

 Y 1 3 9

X
2 1/8 1/24 1/12
4 1/4 1/4 0
6 1/8 1/24 1/12
Marginal distributions of X and Y

xi 2 4 6 yj 1 3 9
f(xi) 1/4 1/2 1/4 g(yj) 1/2 1/3 1/6

 COV (X,Y) = E (XY) – E(X) E(Y)

<Dept. Name> <Faculty Name>

 

<Dept. Name> <Faculty Name>

   COV (X,Y) =12-4(3)=0
Thus, COV (X,Y) =0

  4.

<Dept. Name> <Faculty Name>

 xi 2 5 7 yj 3 4 5
f(xi) 1/2 1/4 1/4 g(yj) 1/3 1/3 1/3

( a ) we have , 𝐽 𝑖𝑗 =𝑓 ( 𝑥 𝑖 ) . 𝑔 ( 𝑦 𝑗 ) where𝑖 , 𝑗=1 ,2 , 3

 

<Dept. Name> <Faculty Name>

 Y 3 4 5 f(xi)
X  
 
2 1/6 1/6 1/6 1/2
5 1/12 1/12 1/12 1/4
7 1/12 1/12 1/12 1/4
g(yj) 1/3 1/3 1/3 1

 
COV (X,Y) = E (XY) – E(X) E(Y)

<Dept. Name> <Faculty Name>

  

COV (X,Y) =16-4(4)=0

Thus, COV (X,Y) =0
 
 c) Z=X+Y

<Dept. Name> <Faculty Name>

 
1/6,1/6,1/6,1/12,1/12,1/12+1/12=1/6,1/12,1/12

T  he probability distribution of Z= X+𝑌 is

Z 5 6 7 8 9 10 11 12
P(Z) 1/6 1/6 1/6 1/12 1/12 1/6 1/12 1/12

w enotethat ∑ P ( z ) =1
 

<Dept. Name> <Faculty Name>

 5.The joint probability distribution of two discrete random variables X and Y is given by f(x,y)= k(2x + y)
where x and y are integers such that 0 x , 0 y
  (a) Find the value of the constant k
(b) Find the marginal probability distributions of X and Y
(c) Show that the random variable X and Y are dependent.
Solution: X = = and Y =
f( x,y) = k(2x+y) and the joint probability distribution table is formed as follows.
X Y 0 1 2 3 sum
0 0 k 2k 3k 6k
 a) We must have 42k = 1
1 2k 3k 4k 5k 14k
k = 1/42
2 4k 5k 6k 7k 22
Sum 6k 9k 12k 15k 42k

<Dept. Name> <Faculty Name>

a) Marginal probability distribution is as follows.
xi 0 1 2 yj 0 1 2 3
f(xi) 6/42 4/42 22/42 g(yj) 6/42 9/42 12/4 15/42
= 1/7 =1/3 =11/21 =1/7 =3/14 =2/7 =5/14

 a) It can be easily seen that f(xi) g(yj)

b) Hence the random variables are dependent.

<Dept. Name> <Faculty Name>

 6. A fair coin is tossed thrice. The random variables X and Y are defined as
follows. X = 0 or 1 according as head or tail occurs on the first toss.
Y = Number of heads
(a) Determine the distribution of X and Y
(b) Determine the joint distribution of X and Y
(c) Obtain the expectations of X,Y and XY. Also find S.Ds of X and Y
(d) Compute Covariance and Correlation of X and Y.
Soln: The sample space S and the association of random variables X
and Y is given by the following table

S HHH HHT HTH HTT THH THT TTH TTT

X 0 0 0 0 1 1 1 1
Y 3 2 2 1 2 1 1 0

 (a) Pro The probability distribution of X and Y is found as follows.

X= = and Y =

<Dept. Name> <Faculty Name>

 (a) probaThe probability distribution of X and Y is found as follows.

X= = and Y =

P(X=0) is 4/8 = 1/2, P( X = 1) is 4/8 = 1/2

P(Y=0) is 1/8 , P( Y = 1) is 3/8
P(Y=2) is 3/8 , P( Y = 3) is 1/8
Thus we have the following probability distribution of X and Y

xi 0 1 yj 0 1 2 3

f(xi) 1/2 1/2 g(yj) 1/8 3/8 3/8 1/8

(b)The joint distribution of X and Y is found by computing

Jij = P( X = xi , Y = yj ) where we have

<Dept. Name> <Faculty Name>

 x1 = 0, x2 = 1 and y1 = 0, y2 = 1 , y3 = 2 , y4 = 3

J 11 = P ( X = 0, Y = 0) = 0
(X = 0 implies that their is a head turn out and Y the total number of heads 0 is impossible)
J 12 = P ( X = 0, Y = 1) = 1/8 correspondings to the outcome HTT

J 13 = P ( X = 0, Y = 2) = 2/8=1/4; outcomes are HHT and HTH

J 14 = P ( X = 0, Y = 3) = 1/8;outcome is HHH

J 21 = P ( X = 1, Y = 0) = 1/8,outcome is TTT

J 22 = P ( X = 1, Y = 1) = 2/8=1/4; outcomes are THT ,TTH

J 23 = P ( X = 1, Y = 2) = 1/8,outcome is THH

J 24 = P ( X = 1, Y = 3) = 0 since the outcome is impossible.

<Dept. Name> <Faculty Name>

 (These values can be written quickly by looking at the table of S , X,Y)
The required joint probability distribution of X and Y is as follows.
 
X Y 0 1 2 3 Sum
0 0 1/8 1/4 1/8 1/2
1 1/8 1/4 1/8 0 1/2
Sum 1/8 3/8 3/8 1/8 1

 
(c) = E(X) = = (0) (1/2) + (1)(1/2) = 1/2
= E(Y) = = (0) (1/8) + (1)(3/8) + 2 (3/8) + 3(1/8) = 12/8 =3/2
E (XY) = = 0 + (0+ 1/4+ 2/8 + 0) = 1/2
  and

<Dept. Name> <Faculty Name>

   = (0 + 1/2 ) – 1/4 = 1/4 = (0+3/8 + 3/2 + 9/8) – ( 9/4) = 3- (9/4) = 3/4
 
Thus = 1/2 and =
c) COV (X,Y) = E (XY) – E(X) E(Y)
=1/2 – 3/4 = -1/4
= = =-

<Dept. Name> <Faculty Name>

 Given X={1,2} and Y={2,3,4}

Marginal distributions of X and Y

Yj 2 3 4
xi 1 2
g(y j) 0.2 0.5 0.3
f(xi) 0.3 0.7

<Dept. Name> <Faculty Name>

<Dept. Name> <Faculty Name>
 Suppose X and Y are independent random variables with the following respective distribution, find the joint
Distribution of X and Y. Also verify that COV(X,Y)=0

xi 1 2 Yj -2 5 8
g(Yj) 0.3 0.5 0.2
f(xi) 0.7 0.3

  Since X and Y are independent, the joint distribution J(x,y) is obtained by using the definition that f(xi) g(yj) )

X Y -2 5 8 f(xi)

1 0.21 0.35 0.14 0.7

2 0.09 0.15 0.06 0.3
g(y j) 0.3 0.5 0.2 1

<Dept. Name> <Faculty Name>

COV
  ( X ,Y )= E ( XY ) − E ( X ) E(Y )
 
E(X) = = (1) (0.7) + (2)(0.3) = 1.3
E(Y) = = (-2) (0.3) + (5)(0.5) + (8) + (0.2) = 3.5 =3/2
E (XY) = = 1(-2)(0.21)+1(5)90.35)+1(8)(0.14)+2(-2)(0.09)+2(5)(0.15)+2(8)(0.06)
E (XY) = 4.55

  4.55 - (1.3)(3.5)=0

<Dept. Name> <Faculty Name>

SAMPLING THEORY

Introduction:
Introduction to sampling distributions, standard error, Type-I and Type-II errors. Test of hypothesis for means. Student’s
t - distribution, Chi - square distribution as a test of goodness of fit.
 
Population:
A large collection of individuals or attributes or numerical data is called a Population or Universe.
Eg: 1) All the students in the college
2) All the tigers in a nation

Sample:
• Finite subset of a population is called Sample.
• The total number of individuals in the sample is called sample size and is denoted by ‘n’
• If n > 30 then sample is said to be large sample otherwise small sample.
• Sampling where a member of the population may be selected more than once is called as
sampling with replacement.
• If a member cannot be chosen more than once is called as sampling without replacement.

<Dept. Name> <Faculty Name>

Sampling:
In our day to day life many a time it becomes necessary to draw some valid and reasonable conclusions concerning a large
mass of individuals or things called population. But to examine every individuals of population is practically impossible.
Therefore we prefer to examine a small section which is randomly selected from the population called as sample. The
process of selecting a sample from the population is called as sampling.
It is essential that this sample selection must be a random selection, so that each member of the population has the same
chance of being included in the sample. Thus, the fundamental assumption underlying theory of sampling is random
sampling.

Objectives of sampling: Sampling aims at gathering the maximum information about the population with the minimum
effort, cost and time.  
 
Parameters and Statics: The statistical constant of population such as Mean, Standard deviation, Correlation etc., are
called parameters.
 
Statistics: The statistical constant drawn from the population are called statistics.
 

<Dept. Name> <Faculty Name>

Sampling Distribution:
Consider all possible samples of size n drawn from the population of size N, we can compute the mean. The means of
these samples will not be identical, if we group these means according to their frequency; the frequency distribution so
obtained is called sampling distribution of means. Likewise we can have sampling distribution of Standard deviation,
mode etc.,
 
Standard Error:
The Standard deviation of sampling distribution is called standard error. It is used to access the difference between the
expected and observed values. Reciprocal of the standard error is called Precision.

Statistical hypothesis:
To reach decision about population on the basis of sample information we make certain assumptions about the
population such assumptions may or may not be true is called Statistical hypothesis or Hypothesis.
 
Testing of Hypothesis:
A process of deciding whether to accept or reject the hypothesis is called testing of hypothesis, the test includes the
method of computing the probability of the observed sample, If this probability is less than some pre-assigned values
then we accept the hypothesis otherwise we reject.
 
<Dept. Name> <Faculty Name>
NULL Hypothesis:
The hypothesis formulated for the purpose of rejecting it under the assumption that it is true is called Null hypothesis
and is denoted by H0.
Example:
To test one procedure is better than other then we assume that “There is no difference between the procedures”
To test whether there is a relation between two variables we take “There is no relation between the variables”

Alternative hypothesis:
Any hypothesis complimentary to NULL hypothesis is called alternative hypothesis and is denoted by H 1 or Ha.

Type-I and Type-II Errors

Type-I Error: Wrong decision to reject the null hypothesis when it is actually true.
Type-II Error: Wrong decision to accept the null hypothesis when it is actually false.
Example:
•In a blood test that shows a patient to have a disease wherein the patient does not have a disease is Type - I
Error where as blood test fails to detect the disease in a patient who really has the disease is Type - II Error

<Dept. Name> <Faculty Name>

 Accepting the hypothesis Rejecting the hypothesis

Hypothesis True Correct decision Wrong decision

( Type-I Error )

Hypothesis False Wrong decision Correct decision

( Type-II Error )

<Dept. Name> <Faculty Name>

Significance level
• 

<Dept. Name> <Faculty Name>

 
• 

<Dept. Name> <Faculty Name>

• 

<Dept. Name> <Faculty Name>

 Right Critical Region Left

One tailed test Two tailed test

The following table will be useful for working problems

Test
5% level 1% level
One-tailed test  -1.645 or 1.645 -2.33 or 2.33 
Two-tailed test -1.96 or 1.96  -2.58 or 2.58 
<Dept. Name> <Faculty Name>
Test of significance of proportions
•Normal distribution is the limiting case of binomial distribution.
•Mean and variance for binomial distribution is respectively np and npq.

Sampling distribution for large sample is normal distribution.

•For the normal distribution standard normal variate is x
z


•Hence the static for test of x  np

z
•significance is npq

where x represents observed number of success in the sample of size n

<Dept. Name> <Faculty Name>

 Thank You
Mrs. Divya K.
Assistant Professor,
Dept. of Mathematics,
ATMECE, Mysuru.

<Dept. Name> <Faculty Name>