 CORRELATION

Whenever two variables 𝑥 and y are so related that an

increase in the one is accompanied by an increase of decrease
in the other , then the variables are said to be correlated –

Ex.
(i) The yield of crop varies with the amount of rainfall .
(ii) Price varies with demand .
(iii) Cost of production varies with the cost of raw materials .

Coefficient of correlation

The number showing the degree or extent to which 𝑥 and y are

related to each other, is called the correlation coefficient , denoted
by 𝑟𝑥𝑦 .
(𝒙𝒊 − 𝒙)(𝒚𝒊 − 𝒚)
𝒓𝒙𝒚 =
{ 𝒙𝒊 − 𝒙)𝟐 { 𝒚𝒊 − 𝒚)𝟐

Where 𝑥 = A.M. of 𝑥 and 𝑦 = A.M. of 𝑦

 ( 𝒙𝒊 ) ( 𝒚𝒊 )
𝒙𝒊 𝒚𝒊 −
𝒓𝒙𝒚 = 𝒏
𝒙𝒊 )𝟐 𝒚𝒊 )𝟐
𝒙𝒊𝟐 − 𝒚𝒊𝟐 −
𝒏 𝒏

If the Standard deviation of 𝑥 and 𝑦 are 𝜎𝑥 and 𝜎𝑦 respectively.

Then
𝒙−𝒙 𝟐 𝒚−𝒚 𝟐
𝝈𝒙 = and 𝝈𝒚 =
𝒏 𝒏

Where n is the number of the value of variable of 𝑥 and 𝑦. Then

𝒙−𝒙 𝒚−𝒚
𝒓𝒙𝒚 = 𝒓 =
𝒏.𝝈𝒙.𝝈𝒚

Co-Variance –

𝑥−𝑥 𝑦 −𝑦
𝑐𝑜𝑣(𝑥, 𝑦) =
𝑛

𝑥𝑦
Or 𝑐𝑜𝑣 𝑥, 𝑦 = − 𝑥𝑦
𝑛

1 ( 𝑥𝑖 ) ( 𝑦𝑖 )
Or Cov(𝑥, 𝑦) = 𝑥𝑖 𝑦𝑖 −
𝑛 𝑛

𝑥−𝑥 𝑦 −𝑦
𝑟 = 𝑟(𝑥, 𝑦) =
𝑛 .𝜎𝑥 .𝜎𝑦

𝑐𝑜𝑣 (𝑥,𝑦) 𝑐𝑜𝑣 (𝑥,𝑦)

Or 𝑟= =
𝜎𝑥 .𝜎𝑦 𝑣𝑎𝑟 (𝑥). 𝑣𝑎𝑟 (𝑦)
We know that

Variance = 𝑆𝑡𝑎𝑛𝑑𝑎𝑟𝑑 𝑑𝑒𝑣𝑖𝑎𝑡𝑜𝑛 2

Where –

𝜎 = Standard deviation

𝜎 2 = Variance

Var(𝑥) = 𝜎𝑥 2 ⟹ 𝜎𝑥 = 𝑣𝑎𝑟(𝑥)

Var(𝑦) = 𝜎𝑦 2 ⟹ 𝜎𝑦 = 𝑣𝑎𝑟(𝑦)

Then we can say that

𝑐𝑜𝑣 (𝑥, 𝑦)
𝑟𝑥𝑦 =
𝑣𝑎𝑟 𝑥 . 𝑣𝑎𝑟 (𝑦)
 Characteristics of 𝒓𝒙𝒚

(i) −1 ≤ 𝑟𝑥𝑦 ≤ 1

(ii) If 𝑟𝑥𝑦 = 0, then 𝑥 and y are not corrected .

(iii) If 𝑟𝑥𝑦 > 0, then 𝑦 increases whenever 𝑥 increases.

(iv) If 𝑟𝑥𝑦 < 0, then 𝑦 decreases whenever 𝑥 increases.

(v) If 0 < 𝑟𝑥𝑦 ≤ 1, there is a positive correlation between 𝑥

and 𝑦.

(vi) If −1 ≤ 𝑟𝑥𝑦 < 0 there is a negative correlation between 𝑥

and 𝑦.

(vii) If 𝑟𝑥𝑦 = −1, there is a perfect negative correlation between

𝑥 and 𝑦.

(viii) If 𝑟𝑥𝑦 = 1, there is a perfect positive correlation between 𝑥

and 𝑦.
Nu. Values of r Degree (or level) of correlation
1. 𝑟 = +1 Perfect Positive Correlation
2. 0.75 ≤ 𝑟 < 1 Positive Correlation of high Degree
3. 0.50 ≤ 𝑟 < 0.75 Positive Correlation of middle or Average Degree
4. 0 < 𝑟 < 0.50 Positive Correlation of low Degree
5. 𝑟=0 No Correlation
6. −0.5 < 𝑟 < 0 Negative Correlation of low Degree
7. −0.75 < 𝑟 ≤ −0.5 Negative Correlation of middle or Average Degree
8. −1 < 𝑟 ≤ −0.75 Negative Correlation of high Degree
9. 𝑟 = −1 Perfect Negative Correlation

Co-efficient of Rank Correlation

6 𝐷𝑖2
R= 1 − where 𝐷𝑖 = 𝑅1 − 𝑅2
𝑛 𝑛 2 −1
 REGRESSION –

I. Lines of Regression

(i) Regression Line of 𝑦 on 𝑥

The equation 𝑦 = 𝑎 + 𝑏𝑥 respesents the line of regression of
𝑦 on 𝑥.
We predict the value of 𝑦 from a given value of 𝑥.

(ii) Regression line of 𝑥 on 𝑦

The line 𝑥 = 𝑎 + 𝑏𝑦 represents the line of regression of 𝑥on
𝑦. we predict the value of 𝑥 from of 𝑥 given value of 𝑦.

II. Equation of line of Regression

(i) 𝑦 on 𝑥 –
𝑦 − 𝑦 = 𝑏𝑦𝑥 𝑥 − 𝑥

𝑐𝑜𝑣(𝑥, 𝑦)
𝑦−𝑦= 𝑥−𝑥
𝜎𝑥 2

𝑟𝜎𝑦
𝑦−𝑦= 𝑥−𝑥
𝜎𝑥
𝑟𝜎𝑦
Where 𝑏𝑦𝑥 =
𝜎𝑥
 (ii) 𝑥 on 𝑦 –

𝑥 − 𝑥 = 𝑏𝑥𝑦 𝑦 − 𝑦

𝑐𝑜𝑣(𝑥, 𝑦)
𝑥−𝑥 = 𝑦−𝑦
𝜎𝑦 2

𝑟𝜎𝑥
𝑥−𝑥 = 𝑦−𝑦
𝜎𝑦

𝑟𝜎𝑥
Where 𝑏𝑥𝑦 =
𝜎𝑦

III. Coefficient of Regression

y on x :
𝑟𝜎𝑦 𝑐𝑜𝑣 (𝑥,𝑦 )
(i) 𝑏𝑦𝑥 = =
𝜎𝑥 𝜎𝑥 2

𝑥𝑖 ( 𝑦𝑖)
𝑥𝑖 𝑦𝑖 −
(ii) 𝑏𝑦𝑥 = 𝑛
( 𝑥 𝑖 )2
𝑥 𝑖2 −
𝑛

𝑛 𝑥𝑖 𝑦𝑖 − 𝑥𝑖 ( 𝑦𝑖 )
(iii) 𝑏𝑦𝑥 =
𝑛 𝑥 𝑖2 −( 𝑥 𝑖 )2
 𝒙 on 𝒚 :

𝑟𝜎𝑥 𝑐𝑜𝑣 (𝑥,𝑦 )

(i) 𝑏𝑥𝑦 = =
𝜎𝑦 𝜎𝑦 2

𝑥𝑖 ( 𝑦𝑖)
𝑥𝑖 𝑦𝑖 −
(ii) 𝑏𝑦𝑥 = 𝑛
( 𝑦 𝑖 )2
𝑦𝑖2 −
𝑛

𝑛 𝑥𝑖 𝑦𝑖 − 𝑥𝑖 ( 𝑦𝑖 )
(iii) 𝑏𝑦𝑥 =
𝑛 𝑦𝑖2 −( 𝑦𝑖 )2

IV. Properties of Regression Coefficients –

As we know that
𝑟𝜎𝑦
Coefficient of Regression y on x is 𝑏𝑦𝑥 = ,
𝜎𝑥

𝑟𝜎𝑥
and Coefficient of Regression 𝑥 on 𝑦 is 𝑏𝑥𝑦 = .
𝜎𝑦

Then

𝑟𝑥𝑦 = 𝑏𝑥𝑦 × 𝑏𝑦𝑥

V. If 𝜃 is the angle between two Regression Lines. Then
1−𝑟 2 𝜎𝑥 .𝜎𝑦
tan 𝜃 = .
𝑟 𝜎𝑥 2 +𝜎𝑦 2

𝜋
(i) If 𝑟 = 0 then 𝜃 = i.e. lines are perpendicular.
2

(ii) If 𝑟 = ±1 then 𝜃 = 00 lines are coincidents.

Note :

I. When 𝑟𝑥𝑦 = ±1, then the two regression lines coincide .

II. 𝑟𝑥𝑦 , 𝑏𝑥𝑦 and 𝑏𝑦𝑥 are of the same sing

III. 𝑟𝑥𝑦 is the G.M. of the 𝑏𝑥𝑦 , 𝑏𝑦𝑥 .

1. If 𝑟 is the correlation coefficient of a bivariate distribution, then

a. 0 < r ≤ 1
b. −1 ≤ 𝑟 ≤ 1
c. 𝑟 ≥ 1
d. −1 < 𝑟 < 1
2. If 𝑥𝑖 = 44, 𝑦𝑖 = 55, 𝑥𝑖 𝑦𝑖 = 256 and 𝑛 = 8, 𝑡ℎ𝑒𝑛 cov 𝑥, 𝑦 =?

a. −3.71 b. −4.61 c. −5.81 d. −2.56

3. Let cov 𝑥, 𝑦 = 10, var (x)= 62.5 and var 𝑦 = 31.36 then 𝑟𝑥𝑦 =?
5 4 3 32
a. b. c. d.
7 5 4 125
4. Let 𝑐𝑜𝑣 𝑥, 𝑦 < 0. It means that :

a. Y increasing whenever 𝑥 increasing

b. Y decreasing whenever 𝑥 increasing
c. X and y both decrease
d. None of these
 5. For the data :

X 4 5 6 8 10
Y 12 10 8 7 5
The coefficient of correlation is

a. −0.39 b. −0.81 c. 0.47 d. 0.63

6. If 𝑟𝑥𝑦 = 0 then

a. There is a perfect correlation between 𝑥 and 𝑦

b. 𝑥 and 𝑦 are not correlated .
c. There is a positive correlation between 𝑥 and 𝑦
d. There ix negative correlation between 𝑥 and 𝑦
 7. The ranks obtained by 10 students in Hindi and English in a
class- test are as follows :

Rank 1 2 3 4 5 6 7 8 9 10
in
Hindi
Rank 3 10 5 1 2 9 4 8 7
in 6
English

The coefficient of correlation between their ranks is

a. 0.15 b. 0.224 c. 0.625 d. none of these

8. If cov 𝑥, 𝑦 = −16.5, 𝑣𝑎𝑟 𝑥 = 2.89 and var 𝑦 = 100 then
coefficient of correlation will be

a. 0.97 b. −0.97 c. ±0.97 d. 1

9. Var (𝑥) is :
a. 𝜎𝑥 b. 𝜎𝑦 c. 𝜎𝑥 2 d. 𝜎 𝑦 2
10. If the perfect correlation between the two variables 𝑥 and 𝑦, then
coefficient of correlation will be
a. 1 b. −1 b. 0 d. ±1
11. If Cov 𝑥, 𝑦 = −8.25, 𝑣𝑎𝑟 𝑥 = 8.25 and var 𝑦 = 8.25, then
coefficient of correlation will be :
a. −8.25 b. ±1 c. 1 d. −1
12. If 𝑥 = 15 , 𝑦 = 36, 𝑥𝑦 = 110; 𝑛 = 5, then the covariance
between 𝑥, 𝑦 will be :
a. 0.4 b. −86 c. 86 d. −0.4
13. The formula of cov (𝑥, 𝑦) is
𝑥, 𝑦 𝑥𝑦 𝑥𝑦 𝑥. 𝑦 𝑥𝑦 𝑥. 𝑦 𝑥. 𝑦 𝑥𝑦
a. 2 − b. − c. − d. +
𝑛 𝑛 𝑛 𝑛 𝑛 𝑛2 𝑛2 𝑛
14. If the two variables are not dependent each other. Then the value
of coefficient of correlation will be 𝑥, 𝑦 :
a. 0 b. 1 c. −1 d. 0. 5
15. If the two variables are correlated with positive correlated of
middle, Then the value of coefficient of correlation will be :
a. 0 < 𝑟 < 0.50
b. 0 < 𝑟 ≤ 0.50
c. 0 ≤ 𝑟 < 0.50
d. 0. 50 < 𝑟 < 0.75
16. Coefficient of correlation lies between :
a. Between − 1 and 1
b. Between 0 and 1
c. between −1 and 0
d. Between 0 and ∞
17. If for two variables 𝑥 and 𝑦, cov 𝑥, 𝑦 = 10, 𝜎𝑥2 = 16, 𝜎𝑦2 = 9.
Then the coefficient of correlation will be :

a. 0.65 b. 0.79 c. 0.83 d. 0.93

18. For two variables 𝑥 and 𝑦 cov, 𝑥, 𝑦 = 8, 𝜎𝑥2 = 9, 𝜎𝑥2 = 16, then
coefficient of correlation is :
2 8 9 2
a. b. c. d.
3 3 2 8 2 9
19. The coefficient of correlation is positive , then for the values
variables , 𝑥, 𝑦 when the value of 𝑥 increasing perfectly, then

a. The value of 𝑦 increasing

b. The value of 𝑦 decreasing
c. The value of 𝑦 remains constant
d. Nothing can be said
20. Regression represents the relation between the following
variables :
a. Only two variables
b. Two or more than two variables
c. In three variables
d. None of these
21. If there is no correlation between two variable, the regression
lines will be :

a. Parallel to the 𝑋 −axis and 𝑦 −axis

b. both of the regression lines will be parallel to each other
c. Both of the regression lines will be mutually will be
mutually coincident
d. none of these
22. The coefficient of correlation is the….. of regression
coefficients :

a. Arithmetic mean
b. Harmonic mean
c. Geometric mean
d. none of these
23. The arithmetic mean of the regression coefficient will be ……..
always coefficient of correlation :

a. Less than and equal b. equal c. Small than d. greater than

24. If 𝑏𝑥𝑦 are the coefficient of regression, then the tangent of the
angle between the regression lines will be :

𝑏𝑦𝑥 .𝑏𝑥𝑦 −1 1−𝑏𝑦𝑥 . 𝑏𝑥𝑦 𝑏𝑦𝑥 .𝑏𝑥 𝑦 +1 1+𝑏𝑦𝑥 . 𝑏𝑥𝑦

a. b. c. d. ,
𝑏𝑥𝑦 +𝑏𝑥𝑦 𝑏𝑦𝑥 +𝑏𝑥𝑦 𝑏𝑦𝑥 −𝑏𝑥𝑦 𝑏𝑦𝑥 −𝑏𝑥𝑦
25. The tangent of angle between the two lines of regression is :

1−𝑟 2 𝜎𝑥 .𝜎𝑦 .
a.
𝑟 𝜎𝑥2 +𝜎𝑦2
𝑟 2 −1 𝜎𝑥 .𝜎𝑦 .
b.
𝑟 𝜎𝑥2 +𝜎𝑦2
𝑟 2 −1 𝜎𝑥 .𝜎𝑦 .
c.
𝑟 𝜎𝑥2 +𝜎𝑦2
d. none of these
 1 1
26. If 𝑏𝑦𝑥 = − and 𝑏𝑥𝑦 = − , then coefficient of correlation will
4 4
be :
1 1 1 1
a. b. ± c. d. −
2 4 4 4
27. The regression line of 𝑥 on 𝑦 is
𝜎𝑥
a. 𝑥−𝑥 =𝑟 (𝑦 − 𝑦)
𝜎𝑦
𝜎𝑦
b. 𝑦−𝑦 =𝑟 (𝑥 − 𝑥 )
𝜎𝑥
𝜎𝑥
c. 𝑥−𝑥 = (𝑦 − 𝑦)
𝑟𝜎𝑦
d. none of these
28. If the regression lines of 𝑥 on 𝑦 and on 𝑥 are 𝑥 = 4𝑦 + 5 and
𝑦 = 𝑘𝑥 + 4, Then

a. 1 ≤ 4𝑘 ≤ 0
b. 0 ≤ 4𝑘 ≤ 1
c. 0 < 4𝑘 < 1
d. 1≤ 4𝑘 < 0
29. If 𝑥 , 𝑦 , 𝑥 2, 𝑦 2 𝑎𝑛𝑑 𝑥𝑦 and number of variable is 𝑛
are given then coefficient 𝑏𝑥𝑦 will be :

𝑥𝑦 −𝑛 𝑥. 𝑦
a.
𝑥 2 −𝑛 𝑥 2
𝑛 𝑥𝑦 −𝑛 𝑥. 𝑦
b.
𝑛 𝑥 2− 𝑥 2
𝑛 𝑥𝑦 − 𝑥. 𝑦
c.
𝑛( 𝑥 2 − 𝑥 2 )
𝑥𝑦 −𝑛 𝑥. 𝑦
d.
𝑥 2 −𝑛 𝑥 2
30. If 𝑥 , 𝑦, 𝑥 2 , 𝑦 2 and 𝑥𝑦 𝑎𝑛𝑑 number of variables is 𝑛
are given then coefficient 𝑏𝑦𝑥 will be

𝑛 𝑥𝑦 − 𝑥. 𝑦 𝑛 𝑥𝑦 − 𝑥. 𝑦 𝑛 𝑥𝑦 − 𝑥. 𝑦
a. b. c. d. none of these
𝑛 𝑥2− 𝑥 2 𝑛 𝑦 2− 𝑦 2 𝑛 𝑥2− 𝑥 2
31. If Cov 𝑥, 𝑦 , 𝜎𝑥2 and 𝜎𝑦2 are given, then coefficient of regression
𝑏𝑥𝑦 will be

𝐶𝑜𝑣 (𝑥,𝑦) 𝐶𝑜𝑣 (𝑥,𝑦 ) 𝐶𝑜𝑣 (𝑥,𝑦)

a. b. c. d. none of these
𝜎𝑥2 𝜎𝑦2 𝜎𝑥 𝜎𝑦
32. If Cov(𝑥, 𝑦 ), 𝜎𝑥2 and 𝜎𝑦2 are given , then coefficient of
regression 𝑏𝑥𝑦 will be :

𝐶𝑜𝑣 (𝑥,𝑦) 𝐶𝑜𝑣 (𝑥,𝑦) 𝐶𝑜𝑣 (𝑥,𝑦 )

a. b. c. d. none of these
𝜎𝑥2 𝜎𝑦2 𝜎𝑥 𝜎𝑦
33. The value of coefficient of regression 𝑏𝑥𝑦 is :

𝜎𝑦 𝜎𝑥 𝜎𝑦 𝜎𝑥
a. 𝑟 b. 𝑟 c. d.
𝜎𝑥 𝜎𝑦 𝑟𝜎𝑥 𝑟 𝜎𝑦
34. The value of coefficient of regression 𝑏𝑥𝑦 is :

𝜎𝑦 𝜎𝑥 𝜎𝑦 𝜎𝑥
a. 𝑟 b. 𝑟 c. d.
𝜎𝑥 𝜎𝑦 𝑟𝜎𝑥 𝑟 𝜎𝑦