CORRELATION

Meaning
a) Correlation studies the relationship between two variables in which change in the value
of one variable causes change in the other variable. It is denoted by letter ‘r’. For e.g.
change in price leads to change in quantity demanded.
b) Correlation studies and measures the direction and intensity of relationship among
variables. It measures co-variation not causation.

Types of Correlation
Correlation is classified into positive and negative correlation.
The correlation is said to be positive when the variables move together in the same
direction. For e.g., sale of Ice cream and temperature move in same direction.
The correlation is said to be negative when the variables move in opposite direction. For
e.g., when you spend more time in studying chances of your failing decline.

Examples of positive correlation are:

1. Price of commodity and amount of supply
2. Increase in height and weight.
3. Age of husband and age of wife.
4. The family income and expenditure on luxury items.

Examples of negative correlation are:

1. Sale of woolen garments and day temperature.
2. Demand of a commodity may go down as a result of rise in prices.
3. Yield of crops and price of crops.

Kinds of correlation
1. Positive and Negative correlation.
2. Linear and non – linear correlation.
3. Simple and multiple correlations.
Positive correlation: When both variables move in the same direction. If one increases,
other also increases and vice-versa.
Negative correlation: - When two variables move in the opposite direction, they are
negatively correlated.
Linear Correlation: - When two variables change in a constant proportion.
Non- linear correlation: - When two variables do not change in the same proportion.
Simple correlation – Relationship between two variables are studied.
Multiple Correction – Relationship between three or more than three variables are
studied.
Degrees of Correlation:

Degree Positive Negative

Perfect +1 -1
Between -0.75 and
High Between + 0.75 and + 1 -1
Between -0.25 & -
Moderate Between + 0.25 and + 0.75 0.75
Between 0 and -
Low Between 0 and + 0.25 0.25
Zero 0 0

1. Perfect Correlation - When values of both variables changes at a constant rate

Types – (a) Perfect positive correlation – when values of both variables changes at a
constant ratio in the same direction correlation coefficient value (r) is + 1
(b) Perfect negative correlation – When values of both the variables change at a constant
ratio in opposite direction. Value of coefficient of correlation is -1
2. Absence of correlation : When there is no relation between the variables r = 0
3. Limited degree correlation : The value of r varies between more than O and less
than 1
Types - a) High : r lies between ± 0.7 & 0.999
b) Moderate = r lies between ± 0.5 and + 0.699
c) Low: r < ± 0.5

Different methods of finding correlation

a) Scatter Diagram
b) Karl Pearson’s coefficient method
c) Rank method / Spearman’s coefficient method
a) Scatter Diagram
i. Scatter diagram offers a graphic expression of the direction and degree of
correlation.
ii. A scatter diagram visually presents the nature of association without giving any
specific numerical value.
b) Karl Pearson’s coefficient of correlation
i. A numerical measure of linear relationship between two variables is given by Karl
Pearson’s coefficient of correlation
ii. Karl Pearson’s coefficient of correlation is a quantitative method of calculating
correlation.
iii. It gives a precise numerical value of the degree of linear relationship between two
variables.
iv. Karl Pearson’s coefficient of correlation is also known as product moment
correlation.
v. A relationship is said to be linear if it can be represented by a straight line.
c) Spearman’s coefficient of Rank Correlation
i. Another measure is Spearman’s coefficient of Rank Correlation
ii. It measures the linear association between ranks assigned to individual items
according to their attributes
iii. Attributes are those variables which cannot be numerically measured such as
intelligence of people, physical appearance, honesty etc.

a) Scatter Diagram
A scatter diagram is a useful technique for visually examining the form of relationship,
without calculating any numerical value. In this technique, the values of the two variables
are plotted as points on a graph paper. The cluster of points, so plotted, is referred to as
a scatter diagram. From a scatter diagram, one can get a fairly good idea of the nature of
relationship. In a scatter diagram the degree of closeness of the scatter points and their
overall direction enable us to examine the relationship. If all the points lie on a line, the
correlation is perfect and is said to be unity. If the scatter points are widely dispersed
around the line, the correlation is low. The correlation is said to be linear if the scatter
points lie near a line or on a line.
Merits of Scatter Diagram
1. Most simplest method.
2. Not affected by size of extreme values.
Demerits
1. Exact degree of correlation cannot be found.

b) Karl Pearson’s Method

Scatter diagram gives a rough idea of the correlation between two variables. It gives no
information about the degree of correlation between the variables. For quantitative
measurement of the degree of relationship between the two variables Karl Pearson has
given the formula

xy
r 
N  y

Here,
r = Coefficient of correlation
x = (X - X)
y = (Y - Y)
N = number of observations
 X = Standard deviation of series X
 Y = Standard deviation of series Y
OR
Actual Mean Method

xy
r 
x 2  y 2
Where x = (X - X)
y = (Y - Y)

Assumed Mean Method

 dx  dy 
dxdy -  
 N 
r
(dx 2 ) (dy 2 )
dx 2  dy 2 
N N
Where dx = X – A
dy = Y – A
A = assumed mean

Illustration 1:
Find Karl Pearson correlation coefficient using actual mean method:

X 3 4 5 6 7 8 9
Y 5 8 9 10 11 15 18
Solution

X Deviation x2 Y Deviation Square Multiple of

Deviation deviations
x= 𝑋 − 𝑋̅ y= 𝑌 − 𝑌̅ (y2) (xy)
3 -3 9 5 -6 36 18
4 -2 4 8 -3 9 6
5 -1 1 9 -2 4 2
6 0 0 10 -1 1 0
7 1 1 11 0 0 0
8 2 4 15 4 16 8
9 3 9 19 8 64 24
∑X = 42 ∑x=0 ∑x =28
2
∑Y = 77 ∑y=0 ∑y =130
2
∑xy=58
N=7 N=7
𝑋̅ = 6

xy
r  ; x= x  ( X - X); y  (Y - Y);
x 2  y 2

Substituting

58 58
r    0.96
28 X 130 60.332

Coefficient of Correlation (r) = + 0.96

It is a situation of high positive correlation

Illustration 2: Short Cut Method

Calculate the coefficient of correlation between X and Y using short cut method

X 5 10 15 20 25
Y 12 20 25 30 35

Square Square Multiple of

Deviation Deviation
X Deviation Y Deviation deviations
(dx=X-A) (dy=Y-A)
(dx2) (dy2) (dxdy)
5 -15 225 12 -13 169 195
15 -5 25 20 -5 25 25
20 (A) 0 0 25 (A) 0 0 0
25 5 25 30 5 25 25
40 20 400 40 15 225 300
N=5 ∑dx=5 ∑dx2=675 N=5 ∑dy=2 ∑dy2=444 ∑dxdy=545

 dx  dy 
dxdy -  
r  N 
(dx) 2 (dy ) 2
dx 2
dy 
2

N N

Substituting

 5 2 
545 -  
r  5  
543
 0.99
(5) 2
(2) 2 544.87
675  444 
5 5

Coefficient of Correlation (r) = + 0.99

It is a situation of high positive correlation

Illustration 3 : Step Deviation Method

Find the correlation between price and demand for commodity X

Price 10 15 20 25 30
Demand 50 45 40 35 30

Solution

Multiple
Square Square
Deviation dx'=dx/C1 Dema Deviation dy'=dy/ C2 of
Price Deviation Deviation
(dx=X-A) C1= 5 nd (dy=Y-A) C2= 5 deviations
(dx’2) (dy’2)
(dx’dy’)
10 -10 -2 4 50 10 2 4 -4
15 -5 -1 1 45 5 1 1 -1
20 0 0 0 40 0 0 0 0
25 5 1 1 35 -5 -1 1 -1
30 10 2 4 30 -10 -2 4 -4
35 15 3 9 25 -15 -3 9 -9
N=6 ∑dx'=3 ∑dx2=19 N=6 ∑dy'=-3 ∑dy2=19 ∑dxdy=-19

 dx'  dy ' 

dx' dy ' -  
r  N 
(dx' ) 2 (dy ' ) 2
dx' 2  dy ' 2 
N N

Substituting

 3 * (3) 
 19 -  
 6 
r  1
(3) 2
(3) 2
19  19 
6 6

Coefficient of Correlation (r) = - 1

It is a situation of perfect negative correlation

Merits of Karl Pearson’s Method

1. Helps to find direction of correlation
2. Most widely used method
Demerits of Karl Pearson’s method
2. Based on large number of assumptions
3. Affected by extreme values

c) Spearman’s Rank Correlation Method

Spearman’s rank correlation was developed by the British psychologist C.E. Spearman.
It is used when the variables cannot be measured exactly. In some situations it may not
be possible to give precise values to some variables. In such cases Spearman Rank
Correlation Coefficient can be used.
Under this method the observations are ranked in ascending or descending order using
numbers and measure the degree of relationship between the ranks instead of actual
numerical values.

6D 2
rk = 1 – 3
N N
rk = Spearman’s rank correlation
D2 = Sum of squares of difference of ranks
N = Number of observation

Rank Correlation Coefficient when the ranks are given:

Illustration 4:
The following are the ranks given by 2 judges to 10 contestants in a beauty contest. Find
the correlation between the two judges.
Judge 1 1 2 3 4 5 6 7 8 9 10

Judge 2 1 4 2 5 3 9 7 10 6 8

Solution:
X Y X-Y=D D2

1 1 -0 0

2 4 -2 4
 3 2 1 1

4 5 -1 1

5 3 2 4

6 9 -3 9

7 7 0 0

8 10 -2 4

9 6 3 9

10 8 2 4

36

6D 2
rk = 1 –
N3  N
6 X 36
=1- = 1-0.219 = 0.781
10 X 99

Rank Correlation Coefficient when Ranks not given

Illustration 5
Find the rank correlation coefficient for the following data of marks obtained by six
students of class XI in First Term and Second Term Examination held in 2014.
Student A B C D E F

First 92 89 87 86 68 67
Term

Second 86 83 91 77 72 70
Term

Solution:
 First Term Second Term R1 R2 R1 - R2= D2
D

92 86 1 2 -1 1

89 83 2 3 1 1

87 91 3 1 2 4

86 77 4 4 0 0

68 72 5 5 0 0

67 70 6 6 0 0

6D 2
rk = 1 – 3
N N
6X 6 36
=1- = 1-
6 6
3
210
= 1- 0.1714 = 0.823

Rank Correlation Coefficient when the Ranks are repeated

In some cases it may be possible that two values in the series acquire same rank. In such
cases the rank to be assigned to individual observations is an average of the ranks which
these two individual observations would have acquired if they were different.
In case of repeated ranks the formula is given as:-
1 1
6(D 2  (m 3  m)  (m 3  m))
rk = 1 – 12 12
N N
3

M = number of items with repeated ranks.

D = Rank differences
N = Number of pairs
Illustration 6:
Use the rank correlation coefficient to determine the judgment between two judges in a
quiz contest. The following are the details of the data of the contest:
Student A B C D E F G

Judge 4 5 7 8 6 3 5
A

Judge 11 9 13 7 13 8 8
B

Solution:

Students Judge Judge R1 R2 R1 - D2

1 2 R2= D

A 4 11 2 5 -3 9

B 5 9 3.5 4 -0.5 0.25

C 7 13 6 6.5 0.5 0.25

D 8 7 7 1 6 36

E 6 13 5 6.5 -1.5 2.25

F 3 8 1 2.5 -1.5 2.25

G 5 8 3.5 2.5 1 1
 D 2

=51

1 1
6(D 2  (m 3  m)  (m 3  m))
rk = 1 – 12 12
N3  N
1 1 1
6(51  (2 3  2)  (2 3  2)  (2 3  2)
=1- 12 12 12
73  7
6(51  1.5) 315
=1- =1 -
7 7
3
336
= 1 -0.9375
= 0.0625

Merits of Spearman’s Rank Correlation

1. Simple and easy to calculate

2. Not affected by extreme values

Demerits of Spearman’s Rank Correlation

1. Not Suitable for grouped data
2. Not based on original values of observations.

Exercises
Very Short Answer Type Questions (One Mark Questions)

1. What is absence of correlation?

2. What is scatter diagram?
3. What does it mean if the correlation between two variables is + 1?
4. What is meant by correlation?
5. List some variables where accurate measurement is difficult.
6. What is negative correlation?
7. Give the meaning of positive correlation.
8. What is the range of simple correlation coefficient?
9. State the type of correlation when two variables change in the same ratio.
10. Give two examples of positive correlation?
11. Mention the principal shortcoming of scatter diagram as a method of estimating
correlation.
12. Give two examples of negative correlation.
13. When is rank correlation method used?
14. Mention the names of different methods for measuring correlation.
15. What is the main demerit of spearman’s rank method?
16. Mention the principal shortcoming of Karl Pearson’s coefficient of correlation.
17. If rxy = 0, then the variables x and y are
a. Linearly related
b. not linearly related
c. Independent
18. The unit of correlation coefficient between height in feet and weight in kilograms is
a. kg / feet
b. percentage
c. non- existent
19. Which method of measuring correlation measures any type of relationship?

SHORT ANSWER QUESTIONS

1. What is meant by correlation? What are the properties of coefficient of correlations?
2. Interpret the values of r as 1, - 1 and 0.
3. Calculate the correlation coefficient between x and y and comment on their
relationship :
X -3 -2 -1 1 2 3

Y 9 4 1 1 4 9

Ans: r=0

4. Calculate the correlation coefficient between x and y and comment on their

relationship :
X 1 3 4 5 7 8

Y 2 6 8 10 14 16

Ans: r= +1

5. Plot the following data as a scatter diagram and comment over the result:
 X 11 10 15 13 10 16 13 8 17 14

Y 6 7 9 9 7 11 9 6 12 11

6. Calculate Karl Pearson’s coefficient of correlation on the following data:

X 15 18 21 24 27 30 36 39 42 48

Y 25 25 27 27 31 33 35 41 41 45

Ans : r= 0.98

7. From the following data, compute the product movement correlation between x and y
:

X series Y series

(i) No. of items 15 15

(ii) Arithmetic mean 25 18

(iii) Square of deviations from

arithmetic mean 136 138

(iv) Summation of products of deviations of X and Y series from their

(Ans. r =
0.89) respective means = 122

8. Number of pairs of observations of X and Y series = 10

X series: Arithmetic average = 65; Standard deviation = 23.33
Y series: Arithmetic average = 66; Standard deviation = 14.9
Summation of products of corresponding deviation of X and Y Series = + 2704
Calculate product moment correlation of x and y series.
(Ans. r = +0.78)

9. Does correlation imply causation?

10. Does zero correlation mean independence?
11. How does rank correlation coefficient differ from Karl Pearson’s Coefficient of
correlation?
LONG ANSWER TYPE QUESTIONS

1. Discuss Karl Pearson’s method of calculating coefficient of correlation. Give its merits
and limitations.

2. In a beauty contest, three judges accorded following ranks to 10 participants :

 Judge I 1 6 5 10 3 2 4 9 7 8

Judge II 3 5 8 4 7 10 2 1 6 9

Judge III 6 4 4 8 1 2 3 10 5 7

Find out by Spearman’s Rank difference method which pair of judges has a common taste
in respect of beauty.
(Ans. rk I & II = - 0.21; rk II & III = -0.29; rk I & III = + 0.64)

3. What are the advantages of Spearman’s rank correlation coefficient over Karl
Pearson’s correlation coefficient? Explain the method of calculating Spearman’s rank
correlation coefficient.

4. Following are the heights and weights of 10 students in a class.

Draw a scatter diagram and indicate whether the correlation is positive or negative.
Height 72 62 63 66 70 75 58 78 72
(in inch)

Weight 65 54 55 61 60 54 50 63 65
(in kg.)

5. Calculate the correlation coefficient of the marks obtained by 10 students in

Accountancy and Economics and interpret it.

Marks (in 50 54 56 59 60 62 61 65 67 71 71 74
Accountancy)

Marks (in 22 25 34 28 26 30 32 30 28 34 36 40
Economics)

(Ans : r = + 0.78)
6. Calculate Spearman’s rank correlation from the following data:

X 10 12 8 15 20 25 40
Y 15 10 6 25 16 12 8
(Ans. rk = + 0.14)

7. Two judges in a beauty competition rank the twelve entries as follows. Calculate rank
correlation coefficient.

X 1 2 3 4 5 6 7 8 9 10 11 12

Y 12 9 6 10 3 5 4 7 8 2 11 1

(Ans r = - 0.45)

8. Calculate the rank coefficient of correlation of the following data:

X: 68 75 90 75 50 62 40 35

Y: 10 12 14 10 10 13 9 8

(Ans. r = + 0.76)

9. When is rank correlation coefficient more precise than simple correlation

coefficient?