SELF INSTRUCTIONAL STUDY MATERIAL

PROGRAMME: M.A. DEVELOPMENT STUDIES

COURSE: STATISTICAL METHODS FOR DEVELOPMENT

RESEARCH (MADVS2005C04)

UNIT III-PART A

CORRELATION

PROF.(Dr). KRISHNAN CHALIL

25, APRIL 2020

1
 U N I T – 3 :C O R R E L A T I O N

After reading this material, the learners are expected to :

 Understand the meaning of correlation and regression
 Learn the different types of correlation
 Understand various measures of correlation., and,
 Explain the regression line and equation

3.1. Introduction
By now we have a clear idea about the behavior of single variables using different measures
of Central tendency and dispersion. Here the data concerned with one variable is called
‗univariate data‘ and this type of analysis is called ‗univariate analysis‘. But, in nature, some
variables are related. For example, there exists some relationships between height of father
and height of son, price of a commodity and amount demanded, the yield of a plant and
manure added, cost of living and wages etc. This is a a case of ‗bivariate data‘ and such
analysis is called as ‗bivariate data analysis. Correlation is one type of bivariate statistics.
Correlation is the relationship between two variables in which the changes in the values of
one variable are followed by changes in the values of the other variable.

3.2. Some Definitions

1. ―When the relationship is of a quantitative nature, the approximate statistical tool for
discovering and measuring the relationship and expressing it in a brief formula is
known as correlation‖—(Craxton and Cowden)
2. ‗correlation is an analysis of the co-variation between two or more variables‖—(A.M
Tuttle)
3. ―Correlation analysis attempts to determine the degree of relationship between
variables‖—(Ya Lun Chou)
4. ―Correlation analysis deals with the association between two or more variables‖—
(Simpson and Kofka)

2
Thus, the association of any two variates is kown as correlation. Correlation is the numerical
measurement showing the degree of correlation between two variables. One variable may be
called ―independent‖ and the other ―dependent‖ variable.

3.3. Utility of Correlation

Correlation coefficient is a useful tool for many statistical purposes. Let us what these are:

 Correlation is very useful for Economists to study relationships between variables.

 It helps in measuring the degree of relationship between the variables
 We can also test the significance of the relationship
 Sampling error can also be calculated., and
 Correlation is the basis for the study of regression

3.4. Correlation and Causation

The measure of correlation is only a measure of co-variation. It is a numerical measurement
of the extent to which correlation can be found between two or more than two variables. It
does not prove causation. Correlation may happen because of several reasons like: (i) due to
pure chance, (ii) both the correlated variables may be influenced by one or more other
variables., and , (iii) both the variables may be mutually influencing each other so that neither
can be designated as the cause and the other the effect. Thus, whether the changes in
variables indicate causation or not, must be decided on other evidence than on the degree of
correlation.

Correlation is best used, therefore, as a suggestive and descriptive piece of analysis, rather
than a technique which gives definitive answers. It is often a preparatory piece of analysis,
which gives some clues to what the data might yield, to be followed more sophisticated
techniques such as regression.

3
3.5. Types of Correlation

Types of
Correlation

On the basis of Degree On the basis of number On the basis of Linearity

of correlation of variables

Linear
Simple
Positive Correlation
correlation
correlation

Partial Non-Linear
Negative correlation correlation
Correlation

Multiple
Correlation

3.5.1. Positive and negative correlation

Whether correlation is positive or negative would depend up on the direction of change of the
variables. Correlation is said to be positive when the values of the two variables move in the
same direction so that an increase in the value of one variable is followed by an increase in
the value of the other variable. Or a decrease in the value of one variable is followed by a
decrease in the value of the other variable. Examples of positive correlation are:
(a) Heights and weights.., (b) amount of rainfall and yield of crop., (c) price and supply of
a commodity

Correlation is said to be negative when the values of the two variables move in the opposite
direction so that an increase in the values of one variable is followed by a decrease in the
value of the other variable. Examples of negative correlation are:
(a) price and demand of a commodity. (b) Sales of woolen garments and temperature.. (c)
Vaccinations and illness: The more that people are vaccinated for a specific illness, the less
that illness occurs.

4
3.5.2. Linear and non-linear correlation

If the amount of change in one variable tends to bear a constant ratio to the amount of change
in other variable, then the correlation is said to be linear. Observe the following two variables
X and Y.

 If we plot these coordinates on a graph, we‘ll get a straight line.

Correlation is non linear, if the amount of change in one variable does not bear a constant
ratio to the amount of change in the other variable. If we plot these coordinates on a graph,
we‘ll get a curve.

5
3.5.3. Simple, Partial & Multiple Correlation

Simple Correlation – When we consider only two variables and check the correlation
between them it is said to be Simple Correlation. For example, radius and circumference of a
circle., Price and quantity demanded

Multiple Correlation – When we consider three or more variables for correlation

simultaneously, it is termed as Multiple Correlation. For example, Price of Cola Drink,
Temperature, Income and Demand for Cola. Another example, when we study the
relationship between the yield of rice per acre and both the amount of rainfall and the amount
of fertilizers used, it is a problem of multiple correlation.

Partial Correlation – When one or more variables are kept constant and the relationship is
studied between others, it is termed as Partial Correlation. For example, If we keep Price of
Cola constant and check the correlation between Temperature and Demand for Cola, it is
termed as Partial Correlation.

6
3.6.Methods of studying Correlation

Methods of
studying
Correlation

Graphical
Algebraic method
method

Karl Pearson's Spearman's Rank Concurrent

Scatter Diagram Correlation Correlation Deviation
Coefficient Coefficient Method

Correlation graph

3.6.1. Scatter Diagram

This is the simplest method of studying correlation between two variables. The two variables
x and y are taken on the X and Y axes of a graph paper. Each pair of x and y value we mark
a dot and we get as many points as the number of pairs of observation. By looking through
the scatter of points, we can form an idea as whether the variables are related or not. If all the
plotted points lie on a straight line rising from the lower left hand corner to the upper right
hand corner, correlation is said to be perfectly positive. If all the plotted points lie on a
straight line falling from the upper left hand corner to the lower right hand corner of the
diagram, correlation is said to be perfectly negative. If all the plotted points fall in a narrow
line and the points are rising from the lower left hand corner to the upper right hand corner of
the diagram, there is degree of positive correlation between variables. If the plotted points
fall in a narrow bank and the points are lying from the upper left hand corner to the right hand
corner, there high degree of negative correlation. If the plotted points lie scattered all over the
diagram, there is no correlation between the two variables.

7
Merits and limitations of Scatter diagram
Merits
It is simple and non-mathematical method of studying correlation between variables.
Making a scatter diagram usually is the first step in understanding the relationship between
two variables.

Limitations
In this method we cannot measure the exact degree of correlation between the variables.

3.6.2. KARL PEARSON’S CO-EFFICIENT OF CORRELATION

The Karl Pearson‘s product-moment correlation coefficient (or simply, the Pearson‘s correlation
coefficient) is a measure of the strength of a linear association between two variables and is
denoted by r or rxy(x and y being the two variables involved).This method of correlation attempts
to draw a line of best fit through the data of two variables, and the value of the Pearson
correlation coefficient, r, indicates how far away all these data points are to this line of best fit.It

8
is a mathematical method for measuring correlation between two variables and was suggested by
Karl Pearson, a British Statistician. It is the most widely used method for measuring correlation.
It is defined as:

Interpretation of ‘r’

The value of the coefficient of correlation will always lie between -1 and +1., i.e., –1 ≤ r ≤ 1.
When r = +1, it means, there is perfect positive correlation between the variables. When r = -1,
there is perfect negative correlation between the variables. When r = 0, there is no relationship
between the two variables. The coefficient correlation describes not only the magnitude of
correlation but also its direction. Thus, +0.8 indicates that correlation is positive because the sign
of r is plus and the degree of correlation is high because the numerical value of r(0.8) is close to
1. If r = -0.4, it indicates that there is low degree of negative correlation because the sigh of r is
negative e and the numerical value of r is less than 0.5.

Assumptions

While calculating the Pearson‘s Correlation Coefficient, we make the following assumptions –

 There is a linear relationship (or any linear component of the relationship) between the
two variables

 We keep Outliers either to a minimum or remove them entirely

9
Properties of the Pearson’s Correlation Coefficient

1. r lies between -1 and +1, or –1 ≤ r ≤ 1, or the numerical value of r cannot exceed one
(unity)
2. The correlation coefficient is independent of the change of origin and scale.
3. Two independent variables are uncorrelated but the converse is not true.

Example 1: calculate correlation coefficient for the following data:

X 2 4 5 6 8 11
Y 18 12 10 8 7 5

Solution:

X Y X2 Y2 XY

2 18 4 324 36
4 12 16 144 48
5 10 25 100 50
6 8 36 64 48
8 7 64 49 56
11 5 121 25 55
∑X =36 ∑Y =60 ∑ X2 =266 ∑ Y2 =706 ∑(XY) =293

Substituting the values in the above formula, we have:

10
r = 6 x 293 – 36 x 60
√6 x 266 - 362 √6x706 - 602

= 1758 - 2160
√1590-1296 √4236 -3600
= -402
17.32 x25.22
= -402
436.81
= -0.920

(Note: there is high degree of negative correlation)

Example 2: Calculate correlation between X and Y

(X) 2 3 4 5 6 7 8
(Y) 4 5 6 12 9 5 4

Solution:
X Y X2 Y2 XY

2 4 4 16 8
3 5 9 25 15
4 6 16 36 24
5 12 25 144 60
6 9 36 81 54
7 5 49 25 35
8 4 64 16 32
2 2
∑X =35 ∑Y =45 ∑ X =203 ∑ Y =343 ∑(XY) =228

r = 7 x 228 – 35 x 45
√7 x 203 - 352 √7x343 - 452

= 1596 - 1575
√1421-1225 √2401 -2025
= 21
14 x 19.39
= -402
436.81
= 0.077

11
Probable error of coefficient of Correlation ‘r’

Probable error of the coefficient of correlation is a statistical measure which measures

reliability and dependability of the value of coefficient of correlation. If probable error is
added to or subtracted from the coefficient of correlation it would give two such limits within
which we can reasonably expect the value of coefficient of correlation to vary. Usually, the
coefficient of correlation is calculated from samples. For different samples drawn from the
same population, the coefficient of correlation may vary. But the numerical value of such
variations is expected to be less than the probable error. The formula for calculating probable
error is:
Probable error of ‗r‘ = 0.6745(1 –r2)
√n
Where, 0.6745 is a constant number., ‗r‘ stands for correlation coefficient and ‗n‘ number of
pairs of observation.

The limits for the population correlation coefficient are:

r = ± P.E (r)

Example: if r = 0.6 and n = 64, find probable error and standard error.

Probable error of ‗r‘ = 0.6745(1 –r2)

√n
= 0.6745 (1 –0.36)
√64
=0.6745 x 0.64
8 = 0.054

Standard error = (1 –r2)

√n

= (1 –0.36)
√64
= 0.08

12
Interpretation of coefficient of correlation on the basis of probable error.
 If the coefficient of correlation is less than its probable error, it is not at all significant
 If the coefficient of correlation is more than six times its probable error, it is
significant
 If the probable error is not much and if the coefficient of correlation is 0.5 or more it
is generally considered to be significant.

Coefficient of Determination

The nature and extent of relationship between two variables are indicated by the coefficient
of correlation. An effective way of interpreting ‗r‘ is by way of coefficient of determination.
The coefficient of determination is defined as the ratio of the explained variance to the total
variance and is denoted by r2 .

Coefficient of Determination = r2 = Explained Variance

Total variance

r2 states what percentage of variations in the dependent variable is explained by the

independent variable.

Coefficient of non-determination is the ratio of the unexplained variation to the total

variation. It is denoted by K2.

K2 = unexplained variance
Total variance
= 1 - Explained Variance
Total variance
= 1 – r2

Another concept related in the context is known as coefficient of alienation and it is defined
as: √1 – r2

13
Check you progress 1:

1. Given the following pairs of value of the variables X and Y: (a) make a
scatter diagram., (b) Do you think that there is any correlation between the
variables x and y? (c) Is it positive or negative? , (d) is it high or low? and, (e)
by graphic inspection draw an estimated line.

X 2 3 5 6 8 9
Y 6 5 7 8 12 11

2. The lengths and weights of a sample of six articles manufactured by a factor

are given here. Find the Pearson‘s correlation coefficient. (Ans: r = 0.97)

Length 3 5 6 7 10 11
(X)
Weight 8 12 11 14 16 17
(Y)

3. Find Karl Pearson‘s coefficient of correlation between the values of X and Y

given here under; Also find probable error . (r=-0.86., PE =.062)

X 46 68 72 75 80 70 93 100
Y 64 50 39 48 12 52 46 30

3.6.3. SPEARMAN’S RANK CORRELATION COEFFICIENT

In 1904, C. Spearman introduced a new method of measuring the correlation between two
variables. Instead of taking the values of the variables he considered the ranks (or order) of
the observations and calculated Pearson‘s coefficient of correlation for the ranks. The
correlation coefficient so obtained is called rank correlation coefficient. This measure is
useful in dealing with qualitative characteristics such as intelligence, beauty, morality,
honesty etc. The formula for spearman‘s rank correlation coefficient is:

Where, rs =spearman rank correlation coefficient

D = differences in ranks between paired items (R1 –R2)
N = number of pairs of observations

Two types of situations may happen here. One is where we are given ranks and the other is
where we are not given any ranks.

14
 (a) When ranks are given:
When actual ranks are given, we can follow the steps as: (i) compute the difference between
two ranks (R1 and R2) and denote it as ‗d‘., (ii) square the ‗d‘ and obtain ∑d2 , and (iii)
substitute the values in the formula.

Example: From the following data, calculate Spearman‘s rank correlation

Rank in 1 2 3 4 5 6 7 8 9 10
Economics
Rank in 4 8 2 3 5 7 6 9 10 1
Statistics

Solution:
R1 R2 d d2 Steps for solution
1 4 -3 9
2 8 -6 36 r = 1 - 6∑d2
3 2 1 1 n(n2-1)
4 3 1 1 =1 – 6(132)
5 5 0 0 10(100-1)
6 7 -1 1 = 1- 0.8
7 6 1 1 =0.2
8 9 -1 1
9 10 -1 1
10 1 9 81 Interpretation: The result indicates that there
Total -- ∑d2 =132 is low positive correlation

(b)When ranks are not given:

In case we are given actual data, we must give them rank. We can assign ranks by taking the
largest value as one or the lowest value as one, next to it give as two and the like.

Example : Find the rank correlation coefficient from the following data:

X 17 13 15 16 6 11 14 9 7 12
Y 36 46 35 24 12 18 27 22 2 8

15
Solution:
X Y Rank X RankY d d2 Solving steps
(R1) (R2) (R1 –R2)
17 36 1 2 -1 1
13 46 5 1 4 16 r = 1 - 6∑d2
15 35 3 3 0 0 n(n2-1)
16 24 2 5 -3 9 =1 – 6(44)
6 12 10 8 2 4 10(100-1)
11 18 7 7 0 0 = 1- 0.267
14 27 4 4 0 0 =0.733

9 22 8 6 2 4
7 2 9 10 -1 1
12 8 6 9 -3 9 Note: correlation is highly
44 positive

Calculation of Spearman’s Rank correlation when equal or

repeated ranks occur

While assigning rank, if two or more items have equal values (i.e., if there occur a tie), they
may be given mid rank. Thus, if two items are on the fifth rank, each may ranked as 5 + 6 /2
= 5.5 and the next item in the order of size would be ranked seventh. When two or more
ranks are equal, the following formula is used for computing rank correlation.

Rank correlation coefficient = 1 – 6 (∑d2 + m3 – m)

12
n(n2-1)

Where, m stands for the number of equal ranks. The term, (m3 – m)
12 is to be added in the
numerator for each group of equal rank both in x and y series.
Example: Calculate the rank correlation coefficient for the following data:
X 68 64 75 50 64 80 75 40 55 64
Y 62 58 68 45 81 60 68 48 50 70

16
Solution:
X Y Rank X (R1) RankY (R2) d d2
68 62 4 5 -1 1
64 58 6 7 -1 1
75 68 2.5 3.5 -1 1
50 45 9 10 -1 1
64 81 6 1 5 25
80 60 1 6 -5 25
75 68 2.5 3.5 -1 1
40 48 10 9 1 1
55 50 8 8 0 0
64 70 6 2 4 16
72

In the above table, one can see that:

75 occurs 2 times, i.e., m = 2 m3 – m = 23 – 2 = 6
64 occurs 3 times, i.e., m = 3 m3 – m = 33 – 3 =24
75 occurs 2 times, i.e., m = 2 m3 – m = 23 – 2 = 6
Total m3 – m = ∑( m3 – m ) =36

Rank correlation coefficient = 1 – 6 (∑d2 + m3 – m)

12
2
n(n -1)
= 1 – 6 (72 + 36)
12
10(100-1)

= 1 – 6 x 75
990
= 0.545

Note: How ranks are assigned when there is repetition? In such cases, rank is
the average of ranks due for all of them if they are different. For example, in
the above problem, highest value is the X series is 80. So, it is given rank 1.
Next two values are the same 75. They are given the average of 2 and 3., i.e.,
2.5. Then 4th rank is given to the next highest value 68. Then 64 occur 3 times.
So each 64 is given the average of ranks 5,6,7., i.e, 6 and so on.

17
Merits and demerits of Rank Correlation
Merits
1. It is easy to compute and understand
2. It is highly useful when the data are of a qualitative nature like intelligence, beauty
etc.
3. When the ranks of different item-values are given, this is the only method for finding
the degree of correlation.
Demerits
1. This method cannot be employed for finding out correlation in a grouped frequency
distribution.
2. It is difficult to calculate rank correlation, if we have more than 30 items of
observation as ranking them requires much labour.
3. Compared to Pearson method, rank correlation is not precise.

Check your progress 2:

1. What is rank correlation coefficient? Find the rank correlation coefficient for the
following data:

X 35 36 40 38 37 39 41 40 36 38
Y 65 72 78 77 76 77 80 79 76 75
2. Two judges in a beauty competition rank the 12 entries as follows. What degree of
agreement is there between the judges?

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

3. Below are given the heights of fathers (X), and those of their sons (Y) in centimeters.
Calculate Spearman‘s rank Correlation coefficient.

X 180 155 170 174 160 172 166 170 170

Y 170 165 180 180 164 169 172 170 174

18
 3.6.4. CONCURRENT DEVIATION METHOD

The calculation of correlation coefficient by this method is based on the direction of change
or variation in the two paired variables. This is denoted by rc and varies between +1 and -1.
It is calculated by the following formula.

Steps in calculating correlation by concurrent deviation method

(i) find out the direction of change of X variable, i.e., as compared with the first value,
whether the second value is increasing or decreasing or is constant. If it is increasing put (+)
sign; if it is decreasing put (-) sign (minus) and if it is constant put zero. Similarly, as
compared to second value find out whether the third value is increasing, decreasing or
constant. Repeat the same process for other values. Denote this column by D x.
(ii) In the same manner as discussed above find out the direction of change of Y variable and
denote this column by Dy.
(iii) Multiply Dx with Dy, and determine the value of c, i.e., the number of positive signs.
r = +√+ (2C-n)/n
(iv) Apply the above formula, i.e., c

Note. The significance of + signs, both ( inside the under root and outside the under root ) is
that we cannot take the under root of minus sign. Therefore, if 2C – n/n is negative, this
negative value of multiplied with the minus sign inside would make it positive and we can
take the under root. But the ultimate result would be negative. If 2C-n is positive, then, of
course, we get a positive n n
value of the coefficient of correlation.

19
Example: the following are the marks obtained a group of 10 students in Economics and
Statistics. Calculate correlation by the method of Concurrent Deviation.
ECO 8 36 98 25 75 82 90 62 65 39
STAT 84 51 91 60 68 62 86 58 53 47

Solution:
Marks in Marks in Dx Dy Dx . Dy
Economics (X) Statistics (Y)
8 84
36 51 + - -
98 91 + + +
25 60 - - +
75 68 + + +
82 62 + - -
90 86 + + +
62 58 - - +
65 53 + - -
39 47 - - +
N=9 N =9 C=6

r = +√+ (2C-n)/n = r = +√+ (2x6-9)/9 = ±√12 – 9/9 = ±√3/9 = 0.58

c c

Merits and Demerits of Concurrent Deviation Method

Merits:
1. It is simple to compute
2. It is easy to understand

Demerits:
1. It is not useful if long-term changes are to be considered.
2. The method does not differentiate between small and big variations.
3. It indicates the direction of change only

20
Check your Progress 3:
1. Calculate the coefficient of concurrent deviation from the following data.(Hints: rc =-
0.5)

X 15 18 23 20 21 28 30 29 50
Y 5 1 15 9 25 20 18 29 24

2. What is the meaning of coefficient of determination?---------------------------------------

--------------------------------------------------------------------------------------------------------
--------------------------------------------------------------------------------------------------------
--------------------------------------------------------------------------------------------------------
---------------------------------------------------------------------
3. Explain the terms a) Probable error, and b)standard error------------------------------------
--------------------------------------------------------------------------------------------------------
--------------------------------------------------------------------------------------------------------
-------------------------------------------------

21