 Definition,

 Karl Pearson’s coefficient of correlation,

 Multiple correlation
 Pharmaceuticals examples
 -:Correlation and Regression:-
Correlation:-If two variables are inter-related in such way that change in one
variable brings about in the another variable ,then this type of relation of
variable known as correlation.
Types of correlation:-
1)Positive, Negative and Zero correlation
2)Linear and Non-linear correlation
3)Simple , Partial and Multiple Correlation.
1)Positive correlation:--
If the value of one variable is increase (decrease),then the value of another
variable also increases (decreases),then the such correlation is called as
positive correlation.
-:Examples:-
1) The longer your hair grows, the more shampoo you will need.
2) As a child grows, so does his clothing size.
3) The more it rains, the more sales for umbrellas go up.
4) As the number of trees cut down increases, the probability
of erosion increases.
5) As the temperature decreases, the speed at which molecules move
decreases.
6) As the amount of moisture increases in an environment, the growth of mold
spores increases.
7) As a student’s study time increases, so does his test average.
2)Negative Correlation:-
If the value of one variable is increase (decrease),then the value of another
variable also decreases (increases), then the such correlation is called as
negative correlation.
Examples:-
1) A student who has many absences, has a decrease in grades.
2) If a train increases speed, the length of time to get to the final point
decreases.
3) If a car decreases speed, travel time to a destination increases.
4) As one increases in age, often one's agility decreases.
5) The more one eats, the less hunger one will have.
6) If a chicken increases in age, the amount of eggs it produces decreases.
3)Zero Correlation:-If the change in the value of one variable ,will not affect the
value of other variable , then the
Correlation is zero.
4)Linear Correlation:-The correlation between two variables is said to be linear
correlation if corresponding to a unit change in the value of one variable ,there
is a constant change in the value of the other variable . The linear correlation
between X 𝑎𝑛𝑑 𝑌 is given by
Y= 𝑎 + 𝑏𝑋 or X = c + dY
5)Non-Linear Correlation:- The correlation between two variables is said to be
non-linear correlation if corresponding to a unit change in the value of one
variable ,the other variable does not change at a constant rate but at a
fluctuating rate.
 -: Method of studying Correlation:-
The correlation is studied by following methods
1)Scatter Diagram
2)Coefficient of Correlation
 -: Coefficient of Correlation:-
1)Karl Pearson's Coefficient of Correlation:-As a measure of intensity or
degree of linear relationship between two variables, Karl Pearson
(1867-1936). developed a formula called Correlation Coefficient.
Let 𝑥1 , 𝑥2 , 𝑥3 , 𝑥4 , … , 𝑥𝑛 be the values n of first variable 𝑋 and let
𝑦1 , 𝑦2 , 𝑦3 , 𝑦4 , … , 𝑦𝑛 be the values n of second variable 𝑌 .Then the Correlation
between variables 𝑋 𝑎𝑛𝑑 𝑌 is denoted by 𝑟 𝑋, 𝑌 , 𝑟𝑋𝑌 or just simply 𝑟 and is
given by
𝑛 σ 𝑋𝑌 − σ 𝑋.σ 𝑌
𝑟 = 𝑟 𝑋, 𝑌 =
𝑛 σ 𝑋2− σ 𝑋 2 . 𝑛 σ 𝑌2− σ 𝑌 2
 Method:
1)Draw the following table
X 𝑌 𝑿𝟐 𝒀𝟐 𝑋𝑌
𝑥1 𝑦1 𝑥12 𝑦12 𝑥1 𝑦1
𝑥2 𝑦2 𝑥22 𝑦22 𝑥2 𝑦2
𝑥3 𝑦3 𝑥32 𝑦32 𝑥3 𝑦3
𝑥4 𝑦4 𝑥42 𝑦42 𝑥4 𝑦4
𝑥5 𝑦5 𝑥52 𝑦52 𝑥5 𝑦5
𝑥6 𝑦6 𝑥62 𝑦62 𝑥6 𝑦6
.. .. . . ..
. . . . .
. . . . .
𝑥𝑛 𝑦𝑛 𝑥𝑛2 𝑦𝑛2 𝑥𝑛 𝑦𝑛

෍𝑿 = ෍𝒀 = ෍ 𝑿𝟐 = ෍ 𝒀𝟐 = ෍ 𝑿𝒀 =
Therefore the correlation between 𝑋 and 𝑌 is given by

𝑛 σ 𝑋𝑌 − σ 𝑋 . σ 𝑌
𝑟 = 𝑟 𝑋, 𝑌 =
𝑛 σ 𝑋2 − σ 𝑋 2 . 𝑛 σ 𝑌2 − σ 𝑌 2

Here 𝑛 is the number of pair observations.

Properties of r:-
1)−1 ≤ 𝑟 ≤ +1
2)If the value of r is negative(r<0) , then there is a negative correlation.
3)If 𝑟 = −1, then there is a perfect negative correlation.
4) 2)If the value of r is positive(r>0) , then there is a positive correlation.
5)If 𝑟 = +1, then there is a positive correlation.
6)If 𝑟 = 0, then there is no correlation .
7)If 𝑟 > 0.8,we say that there exist high degree correlation.
8) If 𝑟 < 0.3,we say that correlation is negligible.
9)Let 𝑟𝑋𝑌 be the coefficient of correlation between the variables X and Y . Let
𝑋−𝐴 𝑌−𝐵
𝑈= and V= .Then the correlation coefficient between U and V is given
ℎ 𝑘
by
ℎ𝑘
𝑟𝑈𝑉 = 𝑟𝑋𝑌
ℎ𝑘
a)If h and k both posses same sign ,then 𝑟𝑈𝑉 = 𝑟𝑋𝑌 .
b)If h and k both posses opposite sign ,then 𝑟𝑈𝑉 = −𝑟𝑋𝑌 .
2) Spearman's Rank Correlation Coefficient:-
The Spearman's Rank Correlation Coefficient(R) is given by
6σ 𝐷 2
𝑅 =1−
𝑛(𝑛2 −1)

Where 𝐷 is the difference between 𝑅1 𝑎𝑛𝑑 𝑅2 . Here 𝑅1 𝑎𝑛𝑑 𝑅2 are the ranks
given by two judges.
Properties of R:-
1)−1 ≤ 𝑅 ≤ +1.
2)If R is positive ,we can say that the two judges have the same thinking .
3)If R is negative ,we can say that the two judges have the opposite thinking .
9)If 𝑋 𝑎𝑛𝑑 𝑌 be two random variables and 𝑎, 𝑏, 𝑐, 𝑑 are any numbers provided
𝑎 ≠ 0, 𝑐 ≠ 0,then
𝑎𝑐
𝑟 𝑎𝑋 + 𝑏, 𝑐𝑌 + 𝑑 = 𝑟 𝑋, 𝑌 .
𝑎𝑐
-:Examples:-
1) Calculate the correlation coefficient for the following heights (in inches) of
fathers (X) and their sons (Y)
X 1 2 3 4 5

Y 1 2 4 5 7

Solution:- First we draw the table

 X 𝑌 𝑿𝟐 𝒀𝟐 𝑋𝑌
𝟏 1 𝟏 𝟏 𝟏
2 2 4 4 4
3 4 9 16 12
4 5 16 25 20
5 7 25 49 35
σ 𝑿 =15 σ 𝒀 =19 σ 𝑿𝟐 =55 σ 𝒀𝟐 =95 σ 𝑿𝒀 =72

Therefore the correlation between 𝑋 and 𝑌 is given by

𝑛 σ 𝑋𝑌 − σ 𝑋.σ 𝑌 5×72 − 15×19

𝑟 = 𝑟 𝑋, 𝑌 = =
𝑛 σ 𝑋2− σ𝑋 2 . 𝑛 σ 𝑌2− σ𝑌 2 5×55 − 15 2 5×95 − 19 2

360−285 75 75 75
= = = = = 0.99
275−225. 475−361 50. 114 5700 75.498
Thus 𝑟 = 0.99.
2)Calculate the correlation coefficient between 𝑋 and 𝑌.
X 10 20 30 40 50 60 70
Y 12 22 32 42 52 62 72
𝑋−10 𝑌−12
Solution: Define 𝑈 = ,𝑉 = .Draw the following table
10 10
X 10 20 30 40 50 60 70
Y 12 22 32 42 52 62 72
U 0 1 2 3 4 5 6
෍ 𝑼 = 𝟐𝟏

V 0 1 2 3 4 5 6
෍ 𝑽 = 𝟐𝟏

𝑈2 0 1 4 9 16 25 36
෍ U 2 = 91

𝑉2 0 1 4 9 16 25 36
෍ V 2 = 91

𝑈𝑉 0 1 4 9 16 25 36
෍ 𝑼𝑽 = 𝟗𝟏
Therefore the correlation between 𝑈 and 𝑉 is given by

𝑛 σ 𝑈𝑉 − σ 𝑈 σ 𝑉
𝑟𝑈𝑉 =
𝑛 σ 𝑈2 − σ 𝑈 2 . 𝑛 σ 𝑉2 − 𝑉 2

7 × 91 − 21 × 21
=
7 × 91 − 21 2
7 × 91 − 21 2
637 − 441 196 196
𝑟𝑈𝑉 = = = =1
637 − 441. 637 − 441 196. 196 196
𝑋−10 𝑌−12
Thus 𝑟𝑈𝑉 = 1. Here 𝑈 = ,𝑉 = ,ℎ = 10 > 0, 𝑘 = 10 > 0.Thus by change
10 10
of scale and origin property, we have
𝑟𝑋𝑌 = 𝑟𝑈𝑉 = 1
 Line of Regression:-
If we plot the data set X and set Y which have both n observations using
scatter diagram ,then we observed that if the variables are related then then
the point are cluster around some cure. If this curve is a straight line ,then it is
called a line of regression and we can say that there is linear regression
between the these two variables X and Y.
We can find the value of variable Y(dependent variable) given the value of
variable X( Independent variable ).Similarly we can find the value of variable
X(dependent variable) given the value of variable Y( Independent variable
).One line is regression line of Y on X and another line is regression line of X on
Y . Regression line Y on X is used to estimate the value of Y for given value of X
and regression line X on Y is used to estimate the value of X for given value of
Y.
 Regression Coefficients:-
Let 𝑥𝑖 , 𝑦𝑖 , 𝑖 = 1,2,3, … , 𝑛 be a bivariate distribution .
1)Let 𝑌 be dependent variable and 𝑋 be independent variable. Then the line
regression of Y on X is given by 𝑌 = 𝑎𝑋 + 𝑏,where a ,b are constant ,𝑎 ≠ 0.Here ‘𝑎’ is
called as a slope of line of regression of Y on X. These slope ‘a’ is called as the
regression coefficient of regression line Y on X. It represent the increment in the
value of dependent variable Y corresponding to a unit change in the value of
independent variable X . It is denoted by 𝑏𝑌𝑋 and is given by
𝜎𝑌 𝑛 σ 𝑋𝑌 − σ 𝑋.σ 𝑌
𝑏𝑌𝑋 = 𝑅𝑒𝑔𝑟𝑒𝑠𝑠𝑖𝑜𝑛 𝑐𝑜𝑒𝑓𝑓𝑖𝑐𝑖𝑒𝑛𝑡 𝑜𝑓 𝑌 𝑜𝑛 𝑋 = 𝑟. =
𝜎𝑋 𝑛 σ 𝑋2− σ 𝑋 2
Where
𝑟 = 𝑐𝑜𝑟𝑟𝑒𝑙𝑎𝑡𝑖𝑜𝑛 𝑐𝑜𝑒𝑓𝑓𝑖𝑐𝑖𝑒𝑛𝑡 𝑏𝑒𝑡𝑤𝑒𝑒𝑛 𝑋 𝑎𝑛𝑑 𝑌.
𝜎𝑋 =Standard deviation of X;𝜎𝑌 =Standard deviation of Y.
Then the regression line Y on X is given by
σ𝑿 σ𝒀
ത ത ത
𝑌 − 𝑌 = 𝑏𝑌𝑋 𝑋 − 𝑋 ,where 𝑋 = ,𝑌 = .
𝑛 𝑛
Method:
1)Draw the following table
X 𝑌 𝑿𝟐 𝑋𝑌
𝑥1 𝑦1 𝑥12 𝑥1 𝑦1
𝑥2 𝑦2 𝑥22 𝑥2 𝑦2
𝑥3 𝑦3 𝑥32 𝑥3 𝑦3
𝑥4 𝑦4 𝑥42 𝑥4 𝑦4
𝑥5 𝑦5 𝑥52 𝑥5 𝑦5
𝑥6 𝑦6 𝑥62 𝑥6 𝑦6
.. .. . ..
. . . .
. . . .
𝑥𝑛 𝑦𝑛 𝑥𝑛2 𝑥𝑛 𝑦𝑛

෍𝑿 = ෍𝒀 = ෍ 𝑿𝟐 = ෍ 𝑿𝒀 =
2) Find the regression coefficient 𝑏𝑌𝑋 ,given by
𝑛 σ 𝑋𝑌 − σ 𝑋.σ 𝑌
𝑏𝑌𝑋 =
𝑛 σ 𝑋2− σ 𝑋 2

3) Find mean 𝑋ത and 𝑌,given

ത by
σ𝑿 σ𝒀
ത
𝑋= ,𝑌 =
𝑛 𝑛
4)Then the regression line Y on X is given by
σ𝑿 σ𝒀
𝑌 − 𝑌ത = 𝑏𝑌𝑋 𝑋 − 𝑋ത ,where 𝑋ത = ,𝑌 =
𝑛 𝑛
 -:Examples:-
1)Find the line of regression 𝑌 on 𝑋 for the following heights (in inches) of fathers (X) and
their sons (Y)

Solution :Draw the table

X 1 2 3 4 5 σ 𝑿 =15

Y 1 2 4 5 7 σ 𝒀 =19

𝑿𝟐 1 4 9 16 25 σ 𝑿𝟐 =55

XY 1 4 12 20 35 σ 𝑿𝒀 =72
The regression coefficient 𝑏𝑌𝑋 ,given by
𝑛 σ 𝑋𝑌 − σ 𝑋.σ 𝑌 5×72 − 15×19 360−285 75 3
𝑏𝑌𝑋 = = = = = = 1.5
𝑛 σ 𝑋2− σ 𝑋 2 5×55 − 15 2 275−225 50 2

mean 𝑋ത and 𝑌ത are given by

σ𝑿 15 σ𝒀 19
ത
𝑋= = = 3, 𝑌 = = = 3.8
𝑛 5 𝑛 5
Therefore ,regression line Y on X is given by
𝑌 − 𝑌ത = 𝑏𝑌𝑋 𝑋 − 𝑋ത
𝑌 − 3.8 = 1.5 𝑋 − 3
𝑌 − 3.8 = 1.5𝑋 − 4.5
𝑌 = 1.5𝑋 − 4.5 + 3.8
𝑌 = 1.5𝑋 − 0.7
∴ 𝑌 = 1.5𝑋 − 0.7 is a line of regression Y on X.
2)Find the line of regression 𝑌 on 𝑋 for the following data

X 1 3 7 9 11
Y 0 4 5 9 10

Solution :Draw the table

X 1 3 7 9 11
෍ 𝑿 = 𝟑𝟏

Y 0 4 5 9 10
෍ 𝒀 = 𝟐𝟖

𝑿𝟐 1 9 49 81 121
෍ 𝑿𝟐 = 𝟐𝟔𝟏

XY 0 12 35 81 110
෍ 𝑿𝒀 = 238
The regression coefficient 𝑏𝑌𝑋 ,given by
𝑛 σ 𝑋𝑌 − σ 𝑋.σ 𝑌 5×238 − 31×28 1190−868 322
𝑏𝑌𝑋 = = = = = 0.94
𝑛 σ 𝑋2− σ 𝑋 2 5×261 − 31 2 1305−961 344

mean 𝑋ത and 𝑌ത are given by

σ𝑿 31 σ𝒀 28
ത
𝑋= = = 6.2, 𝑌 = = = 5.6
𝑛 5 𝑛 5
Therefore ,regression line Y on X is given by
𝑌 − 𝑌ത = 𝑏𝑌𝑋 𝑋 − 𝑋ത
𝑌 − 5.6 = 0.94 𝑋 − 6.2
𝑌 − 5.6 = 0.94𝑋 − 5.828
𝑌 = 0.94𝑋 − 5.828 + 5.6
𝑌 = 0.94𝑋 − 0.228
∴ 𝑌 = 0.94𝑋 − 0.228 is a line of regression Y on X.
2) Let 𝑋 be dependent variable and 𝑌 be independent variable. Then the line
regression of X on Y is given by X = 𝑐𝑌 + 𝑑, where c ,d are constant ,
𝑐 ≠ 0.Here.Here ‘𝑐’ is called as a slope of line of regression of X on Y. These
slope ‘c’ is called as the regression coefficient of regression line X on Y .It
represent the increment in the value of dependent variable X corresponding to
a unit change in the value of independent variable Y . It is denoted by 𝑏𝑋𝑌 and is
given by
𝜎𝑋 𝑛 σ 𝑋𝑌 − σ 𝑋.σ 𝑌
𝑏𝑋𝑌 = 𝑅𝑒𝑔𝑟𝑒𝑠𝑠𝑖𝑜𝑛 𝑐𝑜𝑒𝑓𝑓𝑖𝑐𝑖𝑒𝑛𝑡 𝑜𝑓 𝑋 𝑜𝑛 𝑌 = 𝑟. =
𝜎𝑌 𝑛 σ 𝑌2− σ 𝑌 2
Where
𝑟 = 𝑐𝑜𝑟𝑟𝑒𝑙𝑎𝑡𝑖𝑜𝑛 𝑐𝑜𝑒𝑓𝑓𝑖𝑐𝑖𝑒𝑛𝑡 𝑏𝑒𝑡𝑤𝑒𝑒𝑛 𝑋 𝑎𝑛𝑑 𝑌.
𝜎𝑋 =Standard deviation of X ;𝜎𝑌 =Standard deviation of Y.
Then the regression line X on Y is given by
σ 𝑋. σ𝑌
ത ത ത
𝑋 − 𝑋 = 𝑏𝑋𝑌 𝑌 − 𝑌 ,where 𝑋 = ,𝑌 =
𝑛 𝑛
 Method:
1)Draw the following table
X 𝑌 𝒀𝟐 𝑋𝑌
𝑥1 𝑦1 𝑦12 𝑥1 𝑦1
𝑥2 𝑦2 𝑦22 𝑥2 𝑦2
𝑥3 𝑦3 𝑦32 𝑥3 𝑦3
𝑥4 𝑦4 𝑦42 𝑥4 𝑦4
𝑥5 𝑦5 𝑦52 𝑥5 𝑦5
𝑥6 𝑦6 𝑦62 𝑥6 𝑦6
.. .. . ..
. . . .
. . . .
𝑥𝑛 𝑦𝑛 𝑦𝑛2 𝑥𝑛 𝑦𝑛

෍𝑿 = ෍𝒀 = ෍ 𝒀𝟐 = ෍ 𝑿𝒀 =
2) Find the regression coefficient 𝑏𝑋𝑌 ,given by
𝑛 σ 𝑋𝑌 − σ 𝑋.σ 𝑌
𝑏𝑋𝑌 ==
𝑛 σ 𝑌2− σ 𝑌 2

3) Find mean 𝑋ത and 𝑌,given

ത by
σ𝑿 σ𝒀
ത
𝑋= ,𝑌 =
𝑛 𝑛
4)Then the regression line X on Y is given by
σ 𝑋. σ𝑌
𝑋 − 𝑋ത = 𝑏𝑋𝑌 𝑌 − 𝑌ത ,where 𝑋ത = ,𝑌 =
𝑛 𝑛
 -:Examples:-
1)Find the line of regression 𝑋on 𝑌 for the following heights (in inches) of fathers (X) and
their sons (Y)

Solution :Draw the table

X 1 2 3 4 5 σ 𝑿 =15

Y 1 2 4 5 7 σ 𝒀 =19

𝒀𝟐 1 4 16 25 49
෍ 𝒀𝟐 = 𝟗𝟓

XY 1 4 12 20 35 σ 𝑿𝒀 =72
the regression coefficient 𝑏𝑋𝑌 ,given by
𝑛 σ 𝑋𝑌 − σ 𝑋.σ 𝑌 5×72 − 15×19 360−285 75
𝑏𝑋𝑌 = = = = = 0.52
𝑛 σ 𝑌 2− σ 𝑌 2 5×95 − 19 2 .475−361 144

mean 𝑋ത and 𝑌ത are given by

σ𝑿 15 σ𝒀 19
𝑋ത = 𝑛 = 5 = 3, 𝑌 = 𝑛 = 5 = 3.8
Therefore, the regression line X on Y is given by
𝑋 − 𝑋ത = 𝑏𝑋𝑌 𝑌 − 𝑌ത
𝑋 − 3 = 0.52 𝑌 − 3.8
𝑋 − 3 = 0.52𝑌 − 1.976
𝑋 = 0.52𝑌 − 1.976 + 3
𝑋 = 0.52𝑌 + 1.024
∴ 𝑋 = 0.52𝑌 + 1.024 is a line of regression 𝑋on 𝑌.
2)Find the line of regression 𝑋 on 𝑌 for the following data

X 1 3 7 9 11
Y 0 4 5 9 10

Solution :Draw the table

X 1 3 7 9 11
෍ 𝑿 = 𝟑𝟏

Y 0 4 5 9 10
෍ 𝒀 = 𝟐𝟖

𝒀𝟐 0 16 25 81 100
෍ 𝒀𝟐 = 𝟐𝟐𝟐

XY 0 12 35 81 110
෍ 𝑿𝒀 = 238
the regression coefficient 𝑏𝑋𝑌 ,given by
𝑛 σ 𝑋𝑌 − σ 𝑋.σ 𝑌 5×238 − 31×28 1190−868 322
𝑏𝑋𝑌 = = = = = 0.99
𝑛 σ 𝑌 2− σ 𝑌 2 5×222 − 28 2 1110−784 326

mean 𝑋ത and 𝑌ത are given by

σ𝑿 31 σ𝒀 28
𝑋ത = 𝑛 = 5 = 6.2, 𝑌 = 𝑛 = 5 = 5.6
Therefore, the regression line X on Y is given by
𝑋 − 𝑋ത = 𝑏𝑋𝑌 𝑌 − 𝑌ത
𝑋 − 6.2 = 0.99 𝑌 − 5.6
𝑋 − 6.2 = 0.99𝑌 −5.544
𝑋 = 0.99𝑌 − 5.544 + 6.2
𝑋 = 0.99𝑌 + 0.656
∴ 𝑋 = 0.99𝑌 + 0.656 is a line of regression 𝑋on 𝑌.
Properties of Regression Coefficients:-
1)The sign of regression coefficients 𝑏𝑋𝑌 , 𝑏𝑌𝑋 and r are same.
2 𝑏𝑥𝑦 + 𝑏𝑦𝑥
2 𝑏𝑋𝑌 × 𝑏𝑌𝑋 = 𝑟 and 2
> 𝑟
3)If one of the regression coefficient is greater than one ,then other must be less
than one . Thus product of two regression coefficients is always less than one.
4) Regression coefficients are independent of change of origin but not the scale.
Let X and Y be two variables with S..D. 𝜎𝑋 𝑎𝑛𝑑 𝜎𝑌 respectively.
Let 𝑟𝑋𝑌 be correlation coefficient between X and Y.
𝑏𝑋𝑌 =regression coefficient of X on Y.
𝑏𝑌𝑋 =regression coefficient of Y on X.
𝑋−𝐴 𝑌−𝐵
Let 𝑈 = and V= , h>0 k>0.
ℎ 𝑘
𝑏𝑈𝑉 =regression coefficient of U on V.
𝑏𝑉𝑈 =regression coefficient of V on U.
ℎ 𝑘
Then 𝑏𝑋𝑌 = 𝑏 and 𝑏𝑌𝑋 = 𝑏
𝑘 𝑈𝑉 ℎ 𝑉𝑈
 Properties:-
ത 𝑌ത .
1)The two lines of regression intersect each other in the point 𝑋,
2)If the regression line Y on X is expressed in the form
𝑌 = 𝑎𝑋 + 𝑏.Then the coefficient of X is the regression coefficient 𝑏𝑌𝑋 i.e.
𝑏𝑌𝑋 = 𝑎.
3)If the regression line X on Y is expressed in the form
X= cY + d.Then the coefficient of Y is the regression coefficient 𝑏𝑋𝑌 i.e.
𝑏𝑋𝑌 = 𝑐.
4)The angle between two regression lines is denoted by 𝜃 is given
1−𝑟 2 × 𝜎𝑋 𝜎𝑌 1−𝑟 2 × 𝜎𝑋 .𝜎𝑌
tan 𝜃 = 2 +𝜎 2 ⇒ 𝜃 = tan−1 2 +𝜎 2 .
𝑟 𝜎𝑋 𝑌 𝑟 𝜎𝑋 𝑌
Note:-
𝜋
1)If 𝑟 = 0,then tan 𝜃 = ∞ ⇒ 𝜃 = .
2
Thus we say that two regression lines are perpendicular to each other . Hence
if two variables are uncorrelated.
2)If 𝑟 = ±1,then tan 𝜃 = 0 ⇒ 𝜃 = 0 𝑜𝑟 𝜋.
Thus the two lines are parallel to each other or they coincide with each other
.But since both lines of regression pass through the point 𝑋, ഥ 𝑌ത ,they can not
be parallel .Hence in the case of perfect correlation, positive or negative, the
two lines of regression coincides.
 -:Examples:-
1)Obtain the equation of two lines of regression for the following data. Also
obtain the estimate of X for Y=70.
X 65 66 67 67 68 69 70 72
Y 67 68 65 68 72 72 69 71
Solution:-Let 𝑈 = 𝑋 − 68, 𝑉 = 𝑌 − 69.Then
X Total
65 66 67 67 68 69 70 72
Y
67 68 65 68 72 72 69 71
U -3 -2 -1 -1 0 1 2 4 0
V -2 -1 -4 -1 3 3 0 2 0
𝑈2 9 4 1 1 0 1 4 16 36
𝑉2 4 1 16 1 9 9 0 4 44
UV 6 2 4 1 0 3 0 8 24
 σ8𝑖=1 𝑈𝑖 0 σ8𝑖=1 𝑉𝑖 0
ഥ=
𝑈 = = 0 and 𝑉ത = = =0
8 8 𝑛 8
σ8𝑖=1 𝑈𝑖 𝑉𝑖 24
𝑐𝑜𝑣 𝑈, 𝑉 = ഥ ത
− 𝑈𝑉 = = 3
8 8
σ8𝑖=1 𝑈𝑖2 36
𝜎𝑈2 = ഥ
− 𝑈 2
= = 4.5
8 8
σ8𝑖=1 𝑉𝑖2 44
𝜎𝑉2 = − 𝑉ത 2
= = 5.5
8 8
𝑐𝑜𝑣(𝑈,𝑉) 3
Thus , r 𝑈, 𝑉 = = = 0.603 = 𝑟(𝑋, 𝑌).
𝜎𝑈 𝜎𝑉 4.5 5.5

Since correlation is independent of change of origin, we get

𝑟 = 𝑟 𝑋, 𝑌 = 𝑟 𝑈, 𝑉 = 0.603.
𝑈 = 𝑋 − 68 ⇒ 𝑈ഥ = 𝑋ത − 68 ⇒ 𝑋ത = 68
𝑉 = 𝑌 − 69 ⇒ 𝑉ത = 𝑌ത − 69 ⇒ 𝑌ത = 69.
𝜎𝑋 = 𝜎𝑈 = 𝜎𝑈2 = 4.5 = 2.12 and 𝜎𝑌 = 𝜎𝑉 = 𝜎𝑉2 = 5.5 = 2.35

1) Then the regression line Y on X is given by

𝜎𝑌 2.35
ത ത
𝑌 − 𝑌 = 𝑏𝑌𝑋 𝑋 − 𝑋 ,where 𝑏𝑌𝑋 = r = 0.603 × = 0.668
𝜎𝑋 2.12

∴𝑌 − 69 = 0.668 𝑋 − 68
∴𝑌 − 69 = 0.668𝑋 − 45.424
∴𝑌 = 69 + 0.668𝑋 − 45.424
∴𝑌 = 0.668𝑋 + 23.576 is the regression line Y on X ….(1)
2) Then the regression line X on Y is given by
𝜎𝑋 2.12
ത ത
𝑋 − 𝑋 = 𝑏𝑋𝑌 𝑌 − 𝑌 , where 𝑏𝑋𝑌 = r = 0.603 × = 0.544
𝜎𝑌 2.35

∴ 𝑋 − 68 = 0.544 𝑌 − 69
∴ 𝑋 − 68 = 0.544𝑌 − 37.526
∴ 𝑋 + 68 + 0.544𝑌 − 37.526
∴ 𝑋 = 0.544𝑌 + 30.474 is the regression line X on Y ….(2).
c)To estimate of X for Y=70.Put the Y=70 in equation (2),we get
𝑋 = 0.544 × 70 + 30.474 = 68.554
 Line of Regression Y on X.
73
y = 0.668x + 23.576
72 R² = 0.3636

71

70

69
Y

68

67

66

65

64
64 65 66 67 68 69 70 71 72 73
X
2)In a partially destroyed laboratory ,record of an analysis of correlation data
are given below
Variance of X=9,Regression equation :8𝑋 − 10𝑌 + 66 = 0,40𝑋 − 18𝑌 = 214.
What are 1)Mean values of X and Y.
2)The correlation coefficient between X and Y.
3)Standard deviation of Y ?
Solution:-Let 𝜎𝑋2 = 𝑣𝑎𝑟𝑖𝑎𝑛𝑐𝑒 𝑜𝑓 𝑋 = 9.
Regression equation :8𝑋 − 10𝑌 + 66 = 0,40𝑋 − 18𝑌 = 214.
ത 𝑌ത .Thus
1)We know that both lines of regression passing through the point 𝑋,
8𝑋ത − 10𝑌ത + 66 = 0 ⇒ 8𝑋ത − 10𝑌ത = −66 … (𝑖)
And 40𝑋ത − 18𝑌ത = 214 … (𝑖𝑖)
To solve equation (i) and (ii), 5 × 𝑒𝑞 𝑖 − 𝑒𝑞(𝑖𝑖) ,we get
40𝑋ത − 50𝑌ത = −330
− 40𝑋ത − 18𝑌ത = 214
− + −

−544
0 − 32𝑌ത = −544 ത ത
−32𝑌 = −544 ⇒ 𝑌 = = 17
−32
Put these value 𝑌ത = 17 in equation (i),we get
8𝑋ത − 10𝑌ത = −66 ⇒ 8𝑋ത − 10 × 17 = −66
⇒ 8𝑋ത − 170 = −66
⇒ 8𝑋ത = 170 − 66
104
ത
⇒ 8𝑋 = 104 ⇒ 𝑋 = ത = 13
8
Thus, Mean of 𝑋=𝑋ത = 13 and Mean of 𝑌 = 𝑌ത = 17.
2) Regression equation can be written as
8 66
8𝑋 − 10𝑌 + 66 = 0 ⇒ 𝑌 = 𝑋+
10 10
8 4
∴𝑏𝑌𝑋 =regression coefficient of Y on 𝑋 = =
10 5
18 214
and 40𝑋 − 18𝑌 = 214 ⇒ 𝑋 = 𝑌+
40 40
18 9
∴ 𝑏𝑋𝑌 =regression coefficient of X on Y= = .
40 20
We know that 𝑏𝑋𝑌 × 𝑏𝑌𝑋 = 𝑟 2
4 9 2 2 9 9 3
∴ × =𝑟 ⇒𝑟 = ⇒𝑟=± =± .
5 20 25 25 5
3
Since both the regression confidents are positive ,we have 𝑟 = = 0.6.
5
𝜎𝑋
3)We know that 𝑏𝑋𝑌 = r
𝜎𝑌
𝜎𝑋 9 3 3
Thus 𝑏𝑋𝑌 = r ⇒ = ×
𝜎𝑌 20 5 𝜎𝑌
9 9
⇒ =
20 5𝜎𝑌
1 1
⇒ = ⇒ 20 = 5𝜎𝑌 ⇒ 𝜎𝑌 = 4.
20 5𝜎𝑌

Standard deviation of Y=𝜎𝑌 = 4.

3)Obtain the regression lines for following data
X 6 2 10 4 8
Y 9 11 5 8 7

Solution
X 6 2 10 4 8 Total=30
Y 9 11 5 8 7 40
𝑋2 36 4 100 16 64 220
𝑌2 81 121 25 64 49 340
XY 54 22 50 32 56 214

σ5𝑖=1 𝑥𝑖 30 σ5𝑖=1 𝑦𝑖 40
Here 𝑋ത = = = 6, 𝑌ത = = =8
5 5 5 5
σ5𝑖=1 𝑥𝑖2 220
𝜎𝑋2 = − 𝑋ത 2
= − 6 2
= 44 − 36 = 8
5 5
σ5𝑖=1 𝑦𝑖2 340
𝜎𝑌2 = − 𝑌ത 2
= − 8 2
= 68 − 64 = 4
5 5
 σ5𝑖=1 𝑥𝑖 𝑦𝑖 214
𝑐𝑜𝑣 𝑋, 𝑌 = ത ത
− 𝑋𝑌 = − 48 = 42.8 − 48 = −5.2
5 5
𝐶𝑜𝑣 𝑋,𝑌 −5.2 𝐶𝑜𝑣 𝑋,𝑌 −5.2
𝑏𝑋𝑌 = = = −1.3, 𝑏𝑌𝑋 = = = −0.65
𝜎𝑌2 4 𝜎𝑋2 8
1) Then the regression line Y on X is given by
𝑌 − 𝑌ത = 𝑏𝑌𝑋 𝑋 − 𝑋ത ,
∴𝑌 − 8 = −0.65 𝑋 − 6
∴𝑌 − 8 = −0.65𝑋 + 7.735
∴𝑌 = 8 − 0.65𝑋 + 7.735
∴𝑌 = −0.65𝑋 + 11.9
∴𝑌 = −0.65𝑋 + 11.9 is the regression line Y on X ….(1)
2) Then the regression line X on Y is given by
𝑋 − 𝑋ത = 𝑏𝑋𝑌 𝑌 − 𝑌ത
∴ 𝑋 − 6 = −1.3 𝑌 − 8
∴ 𝑋 − 6 = −1.3𝑌 + 10.4
∴𝑋 = 6 − 1.3𝑌 + 10.4
∴𝑋 = −1.3𝑌 + 16.4
∴ 𝑋 = −1.3𝑌 + 16.4 is the regression line X on Y ….(2).
 Line of Regression Y on X
12
y = -0.65x + 11.9
R² = 0.845
10

6
Y

4
Y

0
0 2 4 6 8 10 12
X
 Standard Error of Estimate or Residual Variance:-
1) The regression line Y on X is given by
𝜎𝑌
ത
𝑌−𝑌 =r 𝑋 − 𝑋ത .The standar𝑑 𝑒𝑟𝑟𝑜𝑟 𝑜𝑓 𝑒𝑠𝑡𝑖𝑚𝑎𝑡𝑒 𝑜𝑓 𝑌, 𝑆𝑌 is given by
𝜎𝑋

𝑆𝑌 = 𝜎𝑌 . 1 − 𝑟 2 .
2) The regression line X on Y is given by
𝜎𝑋
ത ത
𝑋 − 𝑋 = 𝑏𝑋𝑌 𝑌 − 𝑌 , where 𝑏𝑋𝑌 = r .
𝜎𝑌

The standar𝑑 𝑒𝑟𝑟𝑜𝑟 𝑜𝑓 𝑒𝑠𝑡𝑖𝑚𝑎𝑡𝑒 𝑜𝑓 𝑌, 𝑆𝑋 is given by

𝑆𝑋 = 𝜎𝑋 . 1 − 𝑟 2 .
Note:-If 𝑟 = ±1,then 𝑆𝑋 = 𝑆𝑌 = 0,thus each deviation is zero and the two lines
of regression are coincident.
 -:Multiple Correlation:-
If information on two variables like height and weight, income and expenditure,
demand and supply, etc. are available and we want to study the linear relationship
between two variables, correlation coefficient serves our purpose which provides the
strength or degree of linear relationship with direction whether it is positive or
negative. But in biological, physical and social sciences, often data are available on
more than two variables and value of one variable seems to be influenced by two or
more variables. For example, crimes in a city may be influenced by illiteracy,
increased population and unemployment in the city, etc. The production of a crop
may depend upon amount of rainfall, quality of seeds, quantity of fertilizers used
and method of irrigation, etc. Similarly, performance of students in university exam
may depend upon his/her IQ, mother’s qualification, father’s qualification, parents
income, number of hours of studies, etc. Whenever we are interested in studying the
joint effect of two or more variables on a single variable, multiple correlation gives
the solution of our problem. In fact, multiple correlation is the study of combined
influence of two or more variables on a single variable.
Suppose, 𝑋1 , 𝑋2 𝑎𝑛𝑑 𝑋3 are three variables having observations on N individuals
or units. Then multiple correlation coefficient of 𝑋1 , 𝑋2 𝑎𝑛𝑑 𝑋3 is the simple
correlation coefficient between 𝑋1 and the joint effect of 𝑋2 𝑎𝑛𝑑 𝑋3 . It can also
be defined as the correlation between 𝑋1 and its estimate based on 𝑋2 𝑎𝑛𝑑 𝑋3 .
Multiple correlation coefficient is the simple correlation coefficient between
a variable and its estimate.
Let us define a regression equation of 𝑋1 , 𝑋2 𝑎𝑛𝑑 𝑋3 as
𝑋1 = 𝑎 + 𝑏12.3 𝑋2 + 𝑏13.2 𝑋3 …(i)
Without loss of generality we can assume that the variables 𝑋1 , 𝑋2 𝑎𝑛𝑑 𝑋3 have
been measured from their respective means, so that
𝑚𝑒𝑎𝑛(𝑋1 ) = 𝑚𝑒𝑎𝑛 𝑋2 = 𝑚𝑒𝑎𝑛 𝑋3 = 0 … (ii)
By equation (i) ,we get
𝑚𝑒𝑎𝑛(𝑋1 ) = 𝑚𝑒𝑎𝑛 𝑎 + 𝑏12.3 𝑋2 + 𝑏13.2 𝑋3
𝑚𝑒𝑎𝑛(𝑋1 ) = 𝑎 + 𝑏12.3 𝑚𝑒𝑎𝑛(𝑋2 ) + 𝑏13.2 𝑚𝑒𝑎𝑛(𝑋3 )
By equation (ii),we get 𝑎 = 0.
Thus the plane of regression of 𝑋1 , 𝑋2 𝑎𝑛𝑑 𝑋3 becomes
𝑋1 = 𝑏12.3 𝑋2 + 𝑏13.2 𝑋3 …(iii)
The coefficients 𝑏12.3 and 𝑏13.2 are known as the partial regression coefficients
of 𝑋1 on 𝑋2 and of 𝑋1 on 𝑋3 respectively.
𝑒1.23 = 𝑏12.3 𝑋2 + 𝑏13.2 𝑋3 … (𝑖𝑣)
is called the estimate of 𝑋1 as given by the plane of regression (iii) and the
quantity
𝑋1.23 = 𝑋1 − 𝑏12.3 𝑋2 − 𝑏13.2 𝑋3 is called the error of estimate or residual.
2 σ𝑛
𝑖=1 𝑥𝑖 −𝑥ҧ
2 σ𝑛
𝑖=1 𝑥𝑖 −𝑥ҧ 𝑥𝑖 −𝑥ҧ 𝐶𝑜𝑣(𝑋,𝑌)
We know th.at 𝜎 = , 𝐶𝑜𝑣 𝑋, 𝑌 = , 𝑟(𝑋, 𝑌) =
𝑛 𝑛 𝜎𝑥 .𝜎𝑦
 σ𝑁
𝑖=1 𝑋𝑖 −𝑋𝑖
2
Thus, 𝜎𝑖2 = , 𝑖 = 1,2,3.
𝑁
σ𝑁
𝑖=𝑗=1 𝑋𝑖 −𝑋𝑖 𝑋𝑗 −𝑋𝑗
𝐶𝑜𝑣 𝑋𝑖 , 𝑋𝑗 = , i, j = 1,2,3 and i ≠ 𝑗.
𝑁
𝐶𝑜𝑣 𝑋𝑖 ,𝑋𝑗
𝑟𝑖𝑗 = 𝑟 𝑋𝑖 , 𝑋𝑗 = , i, j = 1,2,3 and i ≠ 𝑗.
𝜎𝑖 .𝜎𝑗

1 𝑟23 𝑟12 𝑟23 1 𝑟12

Denote 𝜔11 = , 𝜔12 = 𝑟 1 , 𝜔13 =
𝑟23 1 13 𝑟23 𝑟13
𝜎1 𝜔12 𝜎1 𝜔13
Then 𝑏12.3 = − 𝑎𝑛𝑑 𝑏13.2 = −
𝜎2 𝜔11 𝜎3 𝜔11

Substituting these values in (iii), we get the required equation of the

plane of regression of 𝑋1 , 𝑋2 𝑎𝑛𝑑 𝑋3 as
𝜎1 𝜔12 𝜎1 𝜔13 𝑋1 𝑋2 𝑋3
𝑋1 = 𝑋2 + − 𝑋3 ⇒ 𝜔11 + 𝜔12 + 𝜔13 = 0
𝜎2 𝜔11 𝜎3 𝜔11 𝜎1 𝜎2 𝜎3
 Coefficient of Multiple Correlation. -In a tri-variate distribution in which each
of the variables 𝑋1 , 𝑋2 𝑎𝑛𝑑 𝑋3 has N observations. the multiple correlation
coefficient of 𝑋1 , 𝑋2 𝑎𝑛𝑑 𝑋3 usually denoted by 𝑅1.23 . is the simple correlation
coefficient between 𝑋1 and the joint effect of 𝑋2 and 𝑋3 on 𝑋1 .In other words
𝑅1.23 is the correlation coefficient between 𝑋1 and its estimated value as
given by the plane of regression of 𝑋1 , 𝑋2 𝑎𝑛𝑑 𝑋3 .We know that
𝑒1.23 = 𝑏12.3 𝑋2 + 𝑏13.2 𝑋3 and 𝑋1.23 = 𝑋1 − 𝑏12.3 𝑋2 − 𝑏13.2 𝑋3 .Then
𝐶𝑜𝑣 𝑋1 ,𝑒1.23
𝑅1.23 = , 𝑤ℎ𝑒𝑟𝑒 𝐶𝑜𝑣 𝑋1 , 𝑒1.23 = 𝜎12 − 𝜎1.23
2
= 𝑉𝑎𝑟 𝑒1.23
𝑉𝑎𝑟 𝑋1 .𝑉𝑎𝑟 𝑒1.23
𝜎12 −𝜎1.23
2
Thus 𝑅1.23 =
𝜎12 . 𝜎12 −𝜎1.23
2

2
2 𝜎12 −𝜎1.23
2 . 𝜎12 −𝜎1.23
2 2
𝜎1.23
∴𝑅1.23 = = =1−
𝜎12 . 𝜎12 −𝜎1.23
2 𝜎12 𝜎12
2
𝜎1.23
2
∴1 − 𝑅1.23 = .
𝜎12
 1 𝑟12 𝑟13
2 2 2
Denote 𝜔 = 𝑟21 1 𝑟23 = 1 − 𝑟12 − 𝑟13 − 𝑟23 + 2𝑟12 𝑟13 𝑟23
𝑟31 𝑟32 1
1 𝑟23 2 2 2 𝜔
And 𝜔11 = = 1 − 𝑟23 .Then 𝜎1.23 = 𝜎1
𝑟23 1 𝜔11
2
𝜎1.23 𝜔
2 2
Thus , 1 − 𝑅1.23 = ⇒1− 𝑅1.23 =
𝜎12 𝜔11
𝜔 2 −𝑟 2 −𝑟 2 +2𝑟 𝑟 𝑟
1−𝑟12
2 13 23 12 13 23
⇒ 𝑅1.23 =1− =1− 2
𝜔11 1−𝑟23
2 +𝑟 2 −2𝑟 𝑟 𝑟
𝑟12
2 13 12 13 23
⇒ 𝑅1.23 = 2
1−𝑟23

This formula expresses the multiple correlation coefficient in terms of the

total correlation coefficients between the pairs of variables . Where
 Where
𝑟12 = the total correlation coefficient between variable 𝑋1 and 𝑋2 ,
𝑟23 = the total correlation coefficient between variable 𝑋2 and 𝑋3
𝑟13 = the total correlation coefficient between variable 𝑋1 and 𝑋3
2
Note:- 0 ≤ 𝑅1.23 ≤ 1.
Ex:-Find the regression line of 𝑋1 𝑜𝑛 𝑋2 𝑎𝑛𝑑 𝑋3 and obtain𝑅1.23 , 𝑅2.13 , 𝑅3.12 .
𝑋1 2 5 7 11

𝑋2 3 6 10 12

𝑋3 1 3 6 10

25 31 20
Solution:- 𝑋1 = = 6.25, 𝑋2 = = 7.75, 𝑋3 = =5
4 4 4
 𝑿𝟏 𝑿𝟐 𝑿𝟏 − 𝑋1 𝑿𝟐 − 𝑋2 𝑿𝟏 − 𝑋1 𝟐 𝑿𝟐 − 𝑋2 𝟐 𝑨 𝑩

𝑨 𝑩

2 3 -4.25 -4.75 18.0635 22.5625 20.1875

5 6 -1.25 -1.75 1.5625 3.0625 2.1875

7 10 0.75 2.25 0.5625 5.0625 1.6875

11 12 4.75 4.25 22.5625 18.0625 20.1875

Total=25 31 0 0 42.751 48.75 44.25

 σ4𝑖=1 𝑋𝑖 −𝑋1 2 42.751
𝜎12 = = = 10.68775⇒𝜎1 = 3.269
4 4
σ4𝑖=1 𝑋𝑖 −𝑋2 2 48.75
𝜎22 = = = 12.1875⇒𝜎3 = 3.491
4 4
σ4𝑖=1 𝑋𝑖 − 𝑋1 𝑋𝑗 − 𝑋2 44.25
𝐶𝑜𝑣 𝑋1 , 𝑋2 = = = 11.0625
4 4
𝐶𝑜𝑣 𝑋1 ,𝑋2 11.0625
𝑟12 = 𝑟 𝑋1 , 𝑋2 = = = 0.969
𝜎1 .𝜎2 3.269×3.491
 𝑿𝟏 𝑿𝟑 𝑿𝟏 − 𝑋1 𝑿𝟑 − 𝑋3 𝑿𝟏 − 𝑋1 𝟐 𝑿𝟑 − 𝑋3 𝟐 𝑨 𝑩
𝑩
𝑨

2 1 -4.25 -4 18.0635 16 17

5 3 -1.25 -2 1.5625 4 2.5

7 6 0.75 1 0.5625 1 0.75

11 10 4.75 5 22.5625 25 23.75

Total=25 20 0 0 42.751 46 44
 σ4𝑖=1 𝑋𝑖 −𝑋1 2 42.751
𝜎12 = = = 10.68775⇒𝜎1 = 3.269
4 4
σ4𝑖=1 𝑋𝑖 −𝑋3 2 46
𝜎32 = = = 11.5⇒𝜎3 = 3.361
4 4
σ4𝑖=1 𝑋𝑖 − 𝑋1 𝑋𝑗 − 𝑋3 44
𝐶𝑜𝑣 𝑋1 , 𝑋3 = = = 11.
4 4
𝐶𝑜𝑣 𝑋1 ,𝑋2 11
𝑟13 = 𝑟 𝑋1 , 𝑋2 = = =1
𝜎1 .𝜎2 3.269×3.361
 𝑿𝟐 𝑿𝟑 𝑿𝟐 − 𝑋2 𝑿𝟑 − 𝑋3 𝑿𝟐 − 𝑋2 𝟐 𝑿𝟑 − 𝑋3 𝟐 𝑨 𝑩
𝑩
𝑨

3 1 -4.75 -4 22.5625 16 19

6 3 -1.75 -2 3.0625 4 3.5

10 6 2.25 1 5.0625 1 2.25

12 10 4.25 5 18.0625 25 21.25

Total=31 20 0 0 48.75 46 46
 σ4𝑖=1 𝑋𝑖 −𝑋2 2 48.75
𝜎22 = = = 12.1875⇒𝜎3 = 3.491
4 4
σ4𝑖=1 𝑋𝑖 −𝑋3 2 46
𝜎32 = = = 11.5⇒𝜎3 = 3.361
4 4
σ4𝑖=1 𝑋𝑖 − 𝑋1 𝑋𝑖 − 𝑋3 46
𝐶𝑜𝑣 𝑋1 , 𝑋3 = = = 4.18.
4 4
𝐶𝑜𝑣 𝑋1 ,𝑋2 4.18
𝑟23 = 𝑟 𝑋2 , 𝑋3 = = = 0.356
𝜎1 .𝜎2 3.491×3.361
1 𝑟23 1 0.356
Let 𝜔11 = 𝑟 1
=
0.356 1
= 1 − 0.126736 = 0.873
23
𝑟12 𝑟23 0.969 0.356
, 𝜔12 = 𝑟 1 = = 0.969 − 0.356 = 0.613
13 1 1
1 𝑟12 1 0.969
𝜔13 = = = 1 − 0.344964 = 0.655036
𝑟23 𝑟13 0.356 1
𝜎1 𝜔12 3.269 ×0.613
Then 𝑏12.3 = − =− = −0.6575
𝜎2 𝜔11 3.491×0.873
 𝜎1 𝜔13 3.269×0.655036
 𝑏13.2 =− =− = −0.7297
𝜎3 𝜔11 3.361×0.873

∴The required equation of the plane of regression of 𝑋1 , 𝑋2 𝑎𝑛𝑑 𝑋3 as

𝑋1 𝑋2 𝑋3
𝜔11 + 𝜔12 + 𝜔13 = 0
𝜎1 𝜎2 𝜎3
𝑋1 𝑋2 𝑋3
⇒ × 0.873 + × 0.613 + × 0.655036 = 0
3.269 3.491 3.361
⇒ 2.0.267𝑋1 + 0.1755𝑋2 + 0.1948𝑋3 = 0.
 𝑋1 𝑋𝟐 𝑋𝟑 𝑿𝟐𝟏 𝑿𝟐𝟐 𝑿𝟐𝟑 𝑋1 𝑋𝟐 𝑋1 𝑋𝟑 𝑋𝟐 𝑋𝟑

2 3 1 4 9 1 6 2 3

5 6 3 25 36 9 30 15 18

7 10 6 49 100 36 70 42 60

11 12 10 121 144 100 132 110 120

Total=25 31 20 199 289 146 238 169 201