UNIT 5: STATISTICS AND PROBABILITY

Exercise 10.1
1. Find the correlation coefficient between the two variables under the following conditions
(a) Cov (X, Y) = covariance of the variables X and Y = 18
Variance of X = 16 and the variance of Y = 81
n
Sol : Here, Cov (X, Y) = covariance of the variables X and Y = 18
Var (X) = variance of the variable X = 16
Var (Y) = variance of the variable Y = 81
Cov(X, Y) 18 18 1
 r= = = = = 0.5.
var (X) var (Y) 16 × 81 4 × 9 2
(b) Cov (X, Y) = -16.5, Var. (X) = 2.89 and Var. (Y) = 100
Sol n: Here, Cov (X, Y) = -16.5, Var (X) = 2.89, Var (Y) = 100
Cov(X, Y) -16.5
 r= = = - 0.97.
var (X) var (Y) 2.89 × 100

(c) (X - X )2 = 40, (Y - Y )2 = 63 and (X - X ) (Y - Y ) = 35

Sol n: Here, (X - X )2 = 40, (Y - Y )2 = 63,

(X - X ) (Y - Y ) = 35

 r =
 X  X Y  Y  =
35
= 0.697.
 X  X   Y  Y 
2 2 40 × 63

(d) n = 15,  X = 3.2,  Y = 3.4 and (X - X ) (Y - Y ) = 122

Sol n: Here, n = 15, X = 3.2, Y = 3.4, (X - X ) (Y - Y ) = xy = 122

xy 122
 r= = = 0.75.
nx y 15 × 3.2 × 3.4
(e) n = 10, X = 60, Y = 60, X 2 = 400, Y 2 = 580 and XY = 415

Sol n: Here ,n = 10 X = 60

Y = 60 X2 = 400

Y2 = 580 XY = 415

Regression
*188* Solution Manual to Basic Mathematics

nXY - X . Y
 r =
nX - (X)2 nY2 - (Y2)
2

10 × 415 - 60 × 60
=
10 × 400 - (60)2 × 10 × 380 - (60)2
550
= = 0.59.
20 × 46.9
(f) n = 10, X = 5, Y = 3, X2 = 290, Y2 = 300, XY = 115

Sol n: Here, n = 10, X = 5, Y = 3, X2 = 290, Y2 = 300, XY = 115

XY - n X Y
 r =
X2 - n( X )2 Y2 - n( Y )2
115 - 10 × 5 × 3 115 - 150 35
= = =- = - 0.38.
290 - 10 × 52 300 - 10 × 32 40 210 6.32 × 14.5
(g)
Series X Series Y
No. of pair of observations 10 10
Standard deviation 2.05 2.41
Sum of the squares of 42 58
deviations from their
respective means
Sum of the products of deviations of X and Y from their respective means = 36
n
Sol : Here, n = 10, X = 2.05, Y = 2.41,
–– 2 –– 2 –– ––
(X – X ) = 42, (Y – Y ) = 58 and (X – X )(Y – Y ) = 36

 r=
 X  XY  Y  
36
= 0.73
 X  X   Y  Y  42 58
2 2

2. (a) Calculate Karl Pearson's correlation coefficient between the two variables height (in cm)
and weight (in kg) from the data given below
Height 160 162 165 161 162
Weight 63 62 64 60 61
Sol n: Computation of correlation Coefficient
Height Weight x=x- x y=y- y x2 y2 xy
(x) (y)
160 63 -2 1 4 1 -2
162 62 0 0 0 0 0
165 64 3 2 9 4 6
161 60 -1 -2 1 4 2
162 61 0 -1 0 1 0
x=810 y=310 x2=14 y2=10 xy=6

Correlation
 Statistics and Probability* 189*

310
x 810 y
x = = = 162 and y = = 5 = 62
n 5 n
xy 6
 r= = = 0.51.
x2 y2 14 × 10
(b) Find the correlation coefficient between the two variables X and Y from the
following data.
X 5 7 1 3 4
Y 2 3 4 5 6
Sol n: Computation of Correlation Coefficient
X Y x=X- X y=Y- Y x2 y2 xy

5 2 1 -2 1 4 -2
7 3 3 -1 9 1 -3
1 4 -3 0 9 0 0
3 5 -1 1 1 1 -1
4 6 0 2 0 7 0
X=20 Y=20 x = 0 y = 0 x2= 20 y2=10 xy= - 6

X =
x 20
= = 4, Y =
 y  20 = 4
n 5 n 5

⸫ r =
 xy =
6
= - 0.42.
x y
2 2 20 10

(c) Calculate Karl Pearson’s correlation coefficient between the sales and expenses in
thousand rupees of 5 firms.
Sales: 43 41 36 34 50
Expenses 12 24 15 21 19
n
Sol : Computation of Correlation Coefficient’
Sales (X) Expenses (Y) u = X - 41 v = Y - 15 u2 v2 uv
43 12 2 -3 4 9 -6
41 24 0 9 0 81 0
36 15 -5 0 25 0 0
34 21 -7 6 49 36 - 42
50 19 9 4 81 16 36

u = -1 v=16 u2=159 v 2 =142 uv= -12

Here, 41 and 15 be the assumed means of the series X and Y respectively.

Statistics
*190* Solution Manual to Basic Mathematics

⸫ r =
 uv   u. v
n

n  u   u  n  v   v 
2 2 2 2

5(-12) - (-1)16 - 44
= = =– 0.073
2
5159 - (-1) . 5142 - 16 2
794 . 454
(d) Determine the degree of relationship between the ages of the husbands and their wives
from the following data.
Age of the husband 23 22 24 23 26 27
Age of the wives 20 18 20 21 21 22
Sol n: Computation of Correlation Coefficient
Age of the Age of the x=X- y = Y - x2 y2 xy
husband (X) wife (Y) X Y
22 18 -2.167 -2.33 4.69 5.43 5.05
23 20 -1.167 -0.33 1.36 0.11 0.38
23 21 -1.167 0.67 1.36 0.45 -0.78
24 20 -0.167 -0.33 0.028 0.11 0.05
26 21 1.833 0.67 3.36 0.45 1.23
27 22 2.833 1.67 8.03 2.79 9.76
2 2
x=145 y=122 x =18.83 y =9.34 xy=10.66
x 145 y 122
x = = = 24.167; y = = = 20.33
n 6 n 6
xy 10.66
⸫ r= = = 0.80.
x2 y2 4.34 × 3.05
3. From the following table calculate Karl Pearson's correlation coefficient between the two
variables :
X 6 2 10 4 8
Y 9 11 ? 8 7
Arithmetic means of X and Y series are 6 and 8 respectively.
Sol n: Since y = 8. Let y1 be the missing terms in y series.
9 + 11 + y1 + 8 + 7  35 + y1 
Then, y = 8= 40 = 35 + y1  y 1 = 5
5 5
Computation of Correlation Coefficient
x y x=x- x y=y- y x2 y2 xy

6 9 0 1 0 1 0
2 11 -4 3 16 9 -12
10 5 4 -3 16 9 -12
4 8 -2 0 4 0 0
8 7 2 -1 4 1 -2
x = 0 y = 0 x2=49 2
y =20 xy=-26
xy -26 -26 -26
⸫ r= = = = = - 0.92.
x2 y2 40 × 20 800 28.28

Correlation
 Statistics and Probability* 191*

4. In order to find the correlation coefficient between the two variables X and Y from 12
pair of observations, the following calculations were made:
X = 30, Y = 5, X2 = 670, Y2 = 285, XY = 334
On subsequent verification, it was found that the pair (X = 11, Y = 4) was copied wrong,
the correct value being (X = 10, Y = 14). Find the correct value of correlation coefficient.
Sol n: Here,n = 12, X = 30, Y = 5, X2 = 670, Y2 = 285, XY = 334
wrong pair = (X = 11, Y = 4) and correct pair = (X = 10, Y = 14)
correctedX = 30 + 10 - 11 = 29 corrected Y = 5 - 4 + 14 = 15
2 2 2
correctedX = 670 - 11 + 10 = 649 corrected Y2 = 285 - 42 + 14 2 = 465
correctedXY = 334 - 44 + 140 = 480
nXY - X . Y
⸫ Corrected value of correlation coefficient (rc) =
nX2 - (X)2 nY2 - (Y2)
12 × 430 - 29 × 15
=
12 × 649 - (29)2 × 12 × 465 - (15)2
4725 4775
= = = 0.775.
6947 × 5355 6099.28
Exercise 10.2

1. a) In a drawing competition, two judges have given the following ranks for 8 competitors
S.No. 1 2 3 4 5 6 7 8
Judge A: 3 5 4 7 8 6 1 2
Judge B: 6 4 2 8 7 5 1 3
Sol n: Computation of Rank correlation.
SN Judge A Judge B d = R1 – R2 d2
1 3 6 -3 9
2 5 4 1 1
3 4 2 2 4
4 7 8 -1 1
5 8 7 1 1
6 6 5 1 1
7 1 1 0 0
8 2 3 -1 1
d = 0 d2 = 18
6d 2 6  18 396
⸫ Rank correlation coefficient (R) = 1 – =1 – = = 0.79.
n(n2 – 1) 8(8 2 – 1) 504
b) Find the rank correlation coefficient between the ranks of two judges. b)
Ten items ranked by two experts are presented below
Items A B C D E F G H I J
Expert X 7 5 4 8 6 3 1 2 9 10
Expert Y 6 1 4 3 5 9 8 7 10 2
Compute Spearman's rank coefficient.

Statistics
*192* Solution Manual to Basic Mathematics

Sol n: Computation of rank correlation

Items Expert X (R1) Expert Y (R2) d = R1 – R2 d2
A 7 6 1 1
B 5 1 4 16
C 4 4 0 0
D 8 3 5 25
E 6 5 1 1
F 3 9 -6 36
G 1 8 -7 49
H 2 7 -5 25
I 9 10 -1 1
J 10 2 8 64
d = 0 d2 = 218
6d 2 6  218 –318
⸫ Rank correlation coefficient (R) = 1 – =1 – = = – 0.32.
n(n2 – 1) 10(10 2 – 1) 990
2. a) The IQ of 6 students and their respective marks in a certain examination are presented
below;
Student A B C D E F
IQ 125 110 140 130 90 100
Exam. Marks 82 86 90 75 70 87
Compute Spearman's correlation coefficient between the exam marks and IQ.
n
Sol : Computation of rank correlation
Students IQ Exam Rank of Rank of Exam d = R1 – R2 d2
Marks IQ (R1) marks (R2)
A 125 82 3 4 -1 1
B 110 86 4 3 1 1
C 140 90 1 1 0 0
D 130 75 2 5 -3 9
E 90 70 6 6 0 0
F 100 87 5 2 3 9
d = 0 d2 = 20
“Alternative for ranking”
Students IQ Exam Rank of Rank of Exam d = R1 – R2 d2
Marks IQ (R1) marks (R2)
A 125 82 4 3 1 1
B 110 86 3 4 -1 1
C 140 90 6 6 0 0
D 130 75 5 2 3 9
E 90 70 1 1 0 0
F 100 87 2 5 -3 9
2
d = 0 d = 20
6d 2 6  20 90
⸫ Rank correlation coefficient (R) = 1 – =1 – = = 0.43.
n(n2 – 1) 6(6 2 – 1) 210

Correlation
 Statistics and Probability* 193*

b) The marks obtained by 8 students in Statistics and Accountancy are given below
S.No. 1 2 3 4 5 6 7 8
Marks in Stat: 30 50 25 60 70 80 65 75
Marks in Acc: 50 60 30 40 82 90 70 73
Compute Spearman's correlation coefficient between the marks in Statistics and Accountancy.
Sol n: Computation of rank correlation
SN Marks in Marks in Rank on Rank on Acc. d = R1 – R2 d2
Stat. Acc. Stat. (R1) (R2)
1 30 50 2 3 -1 1
2 50 60 3 4 -1 1
3 25 30 1 1 0 0
4 60 40 4 2 2 4
5 70 82 6 7 -1 1
6 80 90 8 8 0 0
7 65 70 5 5 0 0
8 75 73 7 6 1 1
2
d = 0 d = 8
6d 2 68 456
⸫ Rank correlation coefficient (R) = 1 – =1 – = = 0.905.
n(n2 – 1) 8(8 2 – 1) 504
c) Calculate Spearman rank correlation between advertisement cost (in thousand Rs.) and
sales (in Lakhs Rs.) from the following data
Advt. cost 57 45 72 78 53 63 86 98 59 71
Sales 78 37 41 84 56 85 77 87 70 59
Sol n: Computation of rank correlation
Advt. cost Sales Rank on Advt. Rank on sales d = R1 – R2 d2
cost (R1) (R2)
57 78 8 4 4 16
45 37 10 10 0 0
72 41 4 9 -5 25
78 84 3 3 0 0
53 56 9 8 1 1
63 85 6 2 4 16
86 77 2 5 -3 9
98 87 1 1 0 0
59 70 7 6 1 1
71 59 5 7 -2 4
d = 0 d2 = 72
6d 2 6  72 558
⸫ Rank correlation coefficient (R) = 1 – =1 – = = 0.56.
n(n2 – 1) 10(10 2 – 1) 990
3. a) Calculate Spearman rank correlation for the relationship between the demand and
supply of 6 goods.
Items 1 2 3 4 5 6
Demand 100 150 160 200 160 180
Supply 90 120 130 120 150 160

Statistics
*194* Solution Manual to Basic Mathematics

Sol n: Computation of rank correlation

Item Demand Supply Rankon Rank on Supply d = R1 – R2 d2
Demand (R1) (R2)
1 100 90 6 6 0 0
2 150 120 5 4.5 0.5 0.25
3 160 130 3.5 3 0.5 0.25
4 200 120 1 4.5 - 3.5 12.25
5 160 150 3.5 2 1.5 2.25
6 180 160 2 1 1 1

d = 0 d2 = 16
m1(m12 - 1) m2(m22 - 1)
[
6 d 2 +
12
+
12
]
⸫ Rank correlation coefficient (R) = 1 – 2 (m1= 2, m2 = 2)
n(n – 1)
2(2 2 – 1) 2(2 2 – 1)
[
6 16 +
12
+
12
] 6  17
=1– 2 =1– = 0.51.
6(6 – 1) 210
b) The marks obtained by 8 students in Mathematics and Physics are given below;
S. No. 1 2 3 4 5 6 7 8
Marks in Math 40 60 35 68 70 96 70 84
Marks in Phy 48 62 28 52 85 90 52 73
Find the rank correlation coefficient between the marks in Mathematics and Physics.
n
Sol : Computation of rank correlation
SN Marks in Marks in Rank on Rank on Physics d = R1 – R2 d2
Maths Physics Math (R1) (R2)
1 40 48 7 7 0 0
2 60 62 6 4 2 4
3 35 28 8 8 0 0
4 68 52 5 5.5 -0.5 0.25
5 70 85 3.5 2 1.5 2.25
6 96 90 1 1 0 0
7 70 52 3.5 5.5 -2 4
8 84 73 2 3 -1 1
2
d = 0 d = 11.5

m1(m12 - 1) m2(m22 - 1)
[
6 d2 +
12
+
12
]
⸫ Rank correlation coefficient (R) = 1 – 2 (m1= 2, m2 = 2)
n(n – 1)
2(22 – 1) 2(22 – 1)
[
6 11.5 +
12
+
12
] 6  12.5
=1– =1– = 0.85.
8(82 – 1) 504

Correlation
 Statistics and Probability* 195*

c) The following data gives the scores in psychological test (X) and the arithmetic ability (Y)
of 10 children;
Child A B C D E F G H I J
X: 20 25 60 45 80 25 55 65 25 75
Y: 52 50 55 50 60 70 72 78 80 63
Calculate the correlation coefficient between the psychological test and the arithmetic
ability in terms of their ranks.
Sol n: Computation of rank correlation
Child X Y Rank on X (R1) Rank on Y (R2) d = R1 – R2 d2
A 20 52 10 8 2 4
B 25 50 8 9.5 -1.5 2.25
C 60 55 4 7 -3 9
D 45 50 6 9.5 -3.5 12.25
E 80 60 1 6 -5 25
F 25 70 8 4 4 16
G 55 72 5 3 2 4
H 65 78 3 2 1 1
I 25 80 8 1 7 49
J 75 63 2 5 -3 9
d = 0 d2= 131.5
m1(m12 - 1) m2(m22 - 1)
[
6 d 2 +
12
+
12
]
⸫ Rank correlation coefficient (R) = 1 – 2
n(n – 1)
3(32 – 1) 2(22 – 1)
[
6 131.5 +
12
+
12
] 804
=1– 2 =1– = 0.19.
10(10 – 1) 990
4. The following table presents the ranks of 8 competitors given by three judges in a certain
music contest.
Judge I 4 7 8 1 2 5 6 3
Judge II 5 8 4 7 3 2 1 6
Judge III 6 5 3 1 2 4 8 7
Use the method of rank correlation to examine which pair of judges have the nearest approach
to music.
Sol n: Computation of rank correlation
Let, R1, R 2 and R3 are ranks given by the judges to the 8 competitors
Judge I (R1) Judge II (R2) Judge d12 d23 d13
(d )2 (d )2 (d )2
III(R3) = R1 – R2 12 = R2 – R3 23 = R1 – R3 13
4 5 6 -1 1 -1 1 -2 4
7 8 5 -1 1 3 9 2 4
8 4 3 4 16 1 1 5 25
1 7 1 -6 36 6 36 0 0
2 3 2 -1 1 1 1 0 0
5 2 4 3 9 -2 4 1 1
6 1 8 5 25 -7 49 -2 4
3 6 7 -3 9 -1 1 -4 16
d12 = 0 d212 = 98 d23 = 0 d223 = 102 d13 = 0 d213 = 54

Statistics
*196* Solution Manual to Basic Mathematics

Here, no of observations(n) = 8
⸫ Rank correlation coefficient between judges I and II is given by
6d 212 6  98
R12 = 1 – =1 – = – 0.1667.
n(n2 – 1) 8(8 2 – 1)
⸫ Rank correlation coefficient of judges II and III is given by
6  d 2 236  102
R23 = 1 – n(n 2  1) =1 – = – 0.214.
8(8 2 – 1)
⸫ Rank correlation coefficient between judges I and II is given by
6  d 213 6  54
R13 = 1 – n(n 2  1) = 1 – = 0.3571.
8(8 2 – 1)
⸫ Judges I and III have the nearest approach to music.
5. Spearman rank correlation of marks obtained by 10 students in Mathematics and Computer
was found to be 0.6. It was later discovered that difference in rank in two subjects obtained by
one student was incorrectly taken as 9 instead of 7. Find the correct rank correlation
coefficient.
Sol n: Here, n = 10, R = 0.6, d = ?
6d 212
⸫ R= 1–
n(n2 – 1)

 0.6 = 1 – 6d 212  0.6  990 = 990 – 6d2

10(10 2 – 1)
 6d2= 990 – 594  d2 = 66
Correct value of d 2 = 66 – 92 + 7 2 = 34.
6d 212 6  34 990 – 204
⸫ Corrected rank correlation (R)= 1 – =1– = = 0.79.
n(n2 – 1) 10(10 2 – 1) 990
Hints and Solution of MCQ’s
1. Correlation coefficient between two variable shows change in one variable  change in other.
2. Correlation co-efficient measures degree of relationship and direction to which the variables move.
3. The correlation between two variables may be all of a), b) and c)
4. The correlation co-efficient of two variables is measured by all of a), b) and c)
Cov ( x , y) 0
5. If Cor(x, y) = 0, then r = = =0
Var ( x ).Var ( y) Var ( x ).Var ( y)
6. If the variables x and y are independent then Cov (x, y) = 0
Cov ( x , y) 18 18
7. r= = = = 0.5
Var ( x ) Var ( y) 16. 81 4.9

8. If r is correlation co-efficient, then –1  r  1.

9. The nature of correlation co-efficient can be estimated by all a), b) and c).
10. If the correlation co-efficient between x and y is positive then the correlation co-efficient between –
x and –y is positive.

Correlation
 Statistics and Probability* 197*

11. If the correlation co-efficient between x and y is positive then correlation co-efficient between –x
and y is negative.
12. in a scatter diagram, if the dots have a trend to rise from lower left to the upper right, then the
correlation is positive.
13. If x = 20, y = 25, n = 5 and xy = 150
x y
x = = 4, y = =5
n n

( ) (y – y )
1
 Cov (x, y) =  x– x
n

= (xy – x y – x y + x y )
1
n

= [xy – y x – x y +  x y ]
1
n

= [xy – y x – x y + n x y ]
1
n
1
= [150 – 5 × 20 – 4 × 25 + 5 × 4 × 5]
5
1
= [150 – 100 – 100 + 100]
5
= 10
nxy – x. y
14. r =
nx2 – (x)2. ny2 – (y)2
10 × – 85 – 0.0
=
10 × 125 – 0. 10 × 80 – 0
–850 850
= =– = – 0.85
25 2. 20 2 1000
cor. (xy) 30
15. r=  0.75 =
var(x). var(y) var (x). 64
30
 var(x) = =5  x = 5
0.75 × 8
16. Karl Pearson's correlation is independent of change in origin and scale.
17. The degree of relationship between two variables in terms of their performance is determined by
Spearman's rank correlation co-efficient.
18. Spearman's rank correlation can be used in all cases given in a), b) and c)
19. Spearman's rank correlation (R) is such that – 1  R  1.
20. Since r = 1.03 > 1, the correlation co-efficient r = 1.03 is not possible.
21. The respective rank value assigned to greatest to lowest for 50, 42, 36, 36, 10 respectively are 1, 2,
3.5, 3.5, 5
22. Here, n = 6, d = 0 and d2 = – 12
Since, d2 is always non-negative, the information is wrong.
23. rxy = 0.96  between x and y there is high degree correlation and they move in same direction being
positive co-efficient.

Statistics
Exercise 10.3

1. What is regression? Distinguish between correlation and regression.

Sol n: Regression is returning back to the original position. The analysis which is used to describe
the average relationship between the two variables is known as a simple linear regression
analysis.
The following are some of the differences between the correlation and regression :
(i) Correlation means the relationship between the two variables so that the change in the
value of one variable results change in the value of the other variable. But the regression
means returning back to the average value.
(ii) Cause and effect relationship between the two variables is not needed in case of
correlation but there must be cause and effect relationship between two variables in case
of regression.
(iii) Correlation analysis presents the extent to which the two variables are correlated and also
the direction of their movements. But regression analysis aims to study the nature of the
relationship between the two variables so that we may able to find the value of one
variable when the value of other variable is known.
(iv) Correlation coefficient is independent of change of origin and scale but regression
coefficients are dependent of change of origin but not of scale.
(v) The correlation coefficient between the two variables cannot exceed 1 but one regression
coefficient can exceed 1 and other regression coefficient less than 1 making their product
to be less than or equal to 1.
2. What are the lines of regression? Why are there in general two lines of regression?
n
Sol : The relationship between the two variables in the scatter diagram will concentrate around a
certain curve. If the curve is a straight line, then it is known as the line of regression. If x and
y are any two variables quantities then two regression lines are

Regression line y on x: y - y = b xy (x - x )

Regression line x on y: x - x = b xy(y - y )

where, x and y are means of x and y series respectively such that the regression lines

intersecting at ( x , y )
A line of regression gives the best estimate (in the sense of least square method) of one
variable of a given value of the other variable. So, there are in general two lines of regression.
One is the line of regression of x on y giving the best value of x wherever the value of y is
known and the other is the line of regression of y on x giving the best estimated value of y
wherever the value of x is known.
3. Find the correlation coefficients between the two variables under the following conditions
(if possible)
i) b xy = 1.8 and b yx = 0.35
Sol n: Here,b xy = 1.8 and b yx = 0.35, both are positive
⸫ r= b xy × b yx = 1.8 × 0.35 = 0.79.

Statistics
 Statistics and Probability* 199*

ii) b xy = - 0.24 and byx = - 3.25

Sol n: Here,b xy = - 0.24 and b yx = - 3.25, both are negative
( 0.24)  ( 3.25)
⸫ r = - b xy × b yx = - = -0.88.
iii) the two regression coefficients are 1.36 and - 0.8
Sol n: Here,b xy = 1.36 and b yx = - 0.8,
⸫ r= b xy × b yx = (-0.8) × 1.36 = imaginary
 Correlation coefficient is not possible.
iv) the regression coefficient of x on y is 0.56 and that of y on x is 1.24.
Sol n: The regression coefficient of x on y, b xy = 0.56
The regression coefficient of y on x, b yx = 1.24
⸫ r= b xy byx = 0.56  1.24 = 0.83
4. Find the two regression coefficients from the following results
a)  x = 6,  y = 12, r = 0.8
n
Sol : Here,x = 6, y = 12and r = 0.8
y 12
⸫ Regression coefficient of y on x = byx = r = 0.8 × = 1.6 and
x 6
x 6
Regression coefficient of x on y = bxy= r = 0.8 × = 0.4.
y 12
b)  x = 8,  y = 10, r = - 0.6
n
Sol : Here,x = 8, y = 10 and r = - 0.6
y 10
⸫ Regression coefficient of y on x = byx = r = - 0.6 × = - 0.75
x 8
x 8
and Regression coefficient of x on y = bxy = r = - 0.6 × = -0.48.
y 10
5. From the following pair of regression equations, find the regression coefficients,
correlation coefficients and the means of x and y series.
a) 4x - 5y + 33 = 0; 20x - 9y - 107 = 0
Sol n: The given pair of regression equation are
4x - 5y + 33 = 0 .......................... (i)
20x - 9y - 107 = 0 ...................... (ii)
Since the lines passes through ( x , y ), we have

4 x - 5 y + 33 = 0 .......................... (i)

20 x - 9 y - 107 = 0 ...................... (ii)

Solving (i) and (ii) means of x and y series respectively are
x = 13, y = 17.
From the regression equation 4x - 5y + 33 = 0, we have
4 33 4
y= x+ b yx= .
5 4 5

Regression
*200* Solution Manual to Basic Mathematics

From the regression equation 20x - 9y - 107 = 0, we have

9 107 9
x= y+ b xy= .
20 20 20
4 9
 r= b xy × b yx = × = 0.6.
5 20
b) 3x + 2y - 26 = 0; 6x + y - 31 = 0
n
Sol : The given pair of regression equation are
3x + 2y - 26 = 0 ................. (i)
6x + y - 31 = 0 .................... (ii)
Since the lines passes through ( x , y ), we have

3 x + 2 y - 26 = 0 .......................... (i)

6 x + y - 31 = 0 ...................... (ii)
Solving (i) and (ii), we get
x = 7, y = 7
From the regression equation 3x + 2y - 26 = 0, we have
3 26 3
y=- x+ b yx = -
2 3 2
From the regression equation 6x + y - 31 = 0, we have
1 31 1
x=- y+ b xy = -
6 6 6
-3 -1
 r= b xy × b yx = × = 0.5.
2 6
6. a) The regression coefficient of x on y is 0.32 and that of y on x is 0.73. If the arithmetic
means x and y are 40 and 35 respectively, find out the;
i) correlation coefficient between the variable x and y.
ii) regression equations of y on x and x on y.
n
Sol : The regression coefficient of x on y =bxy = 0.32
The regression coefficient of y on x =byx = 0.73

x = 40, y = 35

i) r= b xy byx = 0.32  0.73 = 0.48

ii) The regression coefficient of x on y is
(x - x ) = bxy (y - y )

 y - 40 = 0.32(x - 35)
 y - 40 = 0.32x – 11.2
 x = 0.32x + 28.8
 The regression equation of y on x is

Statistics
 Statistics and Probability* 201*

y - y = b yx (x - x )

 y - 35 = 0.73(x - 40)
 y - 35 = 0.73x – 29.2
 y = 0.73x + 5.8
b) The regression coefficients of x and y and y on x are 0.84 and 0.32 respectively. If the
arithmetic means of x and y series are 42 and 26 respectively, find two equations of
lines of regression. Estimate the value of y when x = 20 and the value of x when y = 30.
Sol n: Here,b xy = 0.84 and b yx = 0.32, x = 42 and y = 26

 The regression equation y on x is

y - y = b yx (x - x )

 y - 26 = 0.32(x - 42)
 y - 26 = 0.32x - 13.44
 y = 0.32x + 12.56 ............... (i)
 The regression equation of x on y is

x - x = b xy(y - y )

 x - 42 = 0.84(y - 26)
 x - 42 = 0.84y - 21.87
 x = 0.84y + 20.16 ........... (ii)
(i) and (ii) are the equations of two regression line
When x = 20, from (i), y = 0.32 × 20 + 12.56 = 18.96
When y = 30, from (ii), x = 0.84 × 30 + 20.16 = 45.36.
7. a) Find the regression equation of y on x when;
i) x = 28, y = 36 and byx = 0.5

Sol n: Here, x = 28, y = 36 and byx = 0.5

Now, the regression equation of y on x is

y - y = b yx (x - x )

 y - 36 = 0.5(x - 28)
 y - 36 = 0.5x - 14
 y = 0.5x + 22.
ii) x = 15, y = 25, x2 = 55, y 2 = 140, xy = 78, n = 5
Sol n: Here,x = 15, y = 25, x2 = 55, y2 = 140, xy = 78 and n = 5
The regression coefficient y on x is given by
nxy - x y 5 × 78 - 15 × 25
byx = = = 0.3
nx2 - (x)2 5 × 55 - (15)2

Regression
*202* Solution Manual to Basic Mathematics

Means of x and y series are,

x 15 y 25
x = = = 3 and y = = =5
n 5 n 5

y - y = b yx (x - x )

 y - 5 = 0.3(x - 3)
 y - 5 = 0.3x - 0.9
 0.3x - y + 4.1 = 0.
 y = 0.3x + 4.1
b) Find the regression equation of x on y when;
i) x = 30, y = 45 and bxy = 0.32

Sol n: Here, x = 30, y = 45 and bxy = 0.32

 The regression equation of x on y is,

x - x = b xy(y - y )

 x -30 = 0.32(y - 45)

 x - 30 = 0.32y – 14.4
 x = 0.32y + 15.6
ii) Find the regression equation of x on y when
x = 6, y = 11, xy = 306, x2 = 164, y 2 = 574, n = 4

Sol n: Here, x = 6, y = 11, xy = 306, x2 = 164, y2 = 574, n = 4

Now, x = n × x = 4 × 6 = 24 and y = n × y = 4 × 11 = 44
The regression coefficient y on x is given by
nxy - x y 4 × 306 - 24 × 44
b xy = = = 0.47
ny2 - (y)2 4 × 574 - (44)2
 The regression equation of x on y is

x - x = bxy (y - y )

 x - 6 = 0.47(y - 11)
 x - 6 = 0.47y - 5.17
 x = 0.47y + 0.83.
8. a) In a correlation study, the following values were obtained.
X Y
Mean 64 72
Standard Deviation 2.4 3.2
Coefficient of correlation = 0.72
Find the two regression equations that are associated with the above values.

Statistics
 Statistics and Probability* 203*

Sol n: Here, x = 64 , y = 72, x = 2.4, y = 3.2

Coefficient of correlation (r) = 0.72
x 3.2
The regression coefficient of x on y = bxy = r = 0.72  = 0.96
y 2.4
 The regression equation of x on y is
x - x = bxy (y - y )
 x - 64 = 0.54(y - 72)
 x - 74 = 0.54y – 38.88
 x = 0.54y + 25.12
Again, the regression equation of y on x is
y - y = b yx (x - x )
 y - 72 = 0.96(x - 64)
 y - 72 = 0.96x – 61.44
 y = 0.96x + 10.56
b) From the following data of rainfall and production of rice, find the most likely
production corresponding to the rainfall of 40 mm
Rainfall (mm) Production (quintals)
Mean 35 50
Standard Deviation 5 8
Coefficient of correlation = 0.8
n
Sol : Let x and y are the variable quantities representing the rainfall and production
respectively
Then, x = 35, y = 50,x = 5, y = 8, r = 0.8

y 8
The regression coefficient y on x = b yx = r = 0.8 × = 1.28
x 5
 The regression equation of y on x is
y - y = b yx (x - x )

 y - 50 = 1.28(x - 35)
 y - 50 = 1.28x - 44.8
 y = 1.28x + 5.2
When x = 40, y = 1.28 × 40 + 5.2 = 56.4 quintals.
 The most likely production corresponding to the rainfall of 40 mm is 56.4 quintals.
c) Given the following data relating the price and supply, estimate the supply when the
price is 50.
Mean price = 100, Mean supply = 103
Variance of price = 64, Variance of supply = 16
Correlation coefficient between the price and the supply = -0.65

Regression
*204* Solution Manual to Basic Mathematics

Sol n: Let, x and y are the variable quantities representing price and supply
Then, x = 100, y = 103, var (x) = 64, var (y) = 16, bxy = -0.65, r = - 0.65

(x - x )2
var (x) = = 64 x = 64 = 8
n

(x  x) 2
var (y) = n = 16 y = 16 = 4
y 4
Now, b yx = r = - 0.65 × = - 0.325
x 8
 The regression equation of y on x is
y - y = b yx (x - x )

 y - 103 = -0.325(x - 100)

 y - 103 = -0.325x + 32.5
 y = -0.325x + 135.5
When x = 50, y = -0.325 × 50 + 135.5 = 119.25.
 The supply when price is 50 is 119.25 quintals.
9. Find the regression equation of x on y from the following data.
x 5 9 13 17 21
y 3 8 13 18 23
Estimate the value of x when y = 18
n
Sol : Computation of Regression equation y on x:
x y u = x - 13 v = y - 13 v2 uv
5 3 -8 -10 100 80
9 8 -4 -5 25 20
13 13 0 0 0 0
17 18 4 5 25 20
21 23 8 10 100 80
u = 0 v = 0 v 2 = 250 uv = 200
Here, n = 5, u = 0, v = 0, v2 = 250, uv = 200, a = 13, b = 13
u
x =a+ = 13 + 0 = 13
n
v
y =b+ = 13 + 0 = 13
n
nuv - u v 5 × 200 - 0 67
bxy = 2 2 = =- = - 1.34
nv - (v) 5 × 250 - 0 50
 The regression equation of x on y is
x - x = b xy (y - y )
4
 x - 13 = (y - 13)
5

Statistics
 Statistics and Probability* 205*

 x - 13 = 0.8y - 10.4
 x = 0.8y + 2.6.
When y = 18, x = 0.8 × 18 + 2.6= 17.
10. Find the regression of y on x from the following data.
x 2 4 5 6 8 11
y 18 12 10 8 7 5
Estimate the value of x when y = 12.
n
Sol : Computation of Regression equation x on y:
x y u=x-6 v = y - 10 v2 uv
2 18 -4 8 16 -32
4 12 -2 2 4 -4
5 10 -1 0 1 0
6 8 0 -2 0 0
8 7 2 -3 4 -6
11 5 5 -5 25 -25
u = 0 v = 0 v 2 = 50 uv = -67
Here, n = 6, u = 0, v = 0, v2 = 50, uv = -67, a = 6, b = 10
u
x =a+ =6+0=6
n
v
y =b+ = 10 + 0 = 10
n
nuv - u v 6 × (-67) - 0 67
byx = = =- = -1.34
nu2 - (u)2 6 × 50 - 0 50
 The regression equation of y on x is
y - y = b yx (x - x )
67
 y - 10 = - (x - 6)
50
 y = -1.34x + 8.04 + 10
 y = - 1.34x + 18.04.
When x = 12, y = - 1.34× 12 + 18.04 = 1.96

Hint and Solution of MCQ's

1. Regression analysis studies the nature of relationship between the variables.
2. There are two lines of regressions
3. When the lines of regression interest a right angle, there is no correlation.
4. Correlation coefficients between two variables is the GM between two regression co-efficient.
5. From the relation r2 = bxy. byx i.e. r =  bxy. byx ,
r will be positive so bxy and byx both positive.
6. The range of bxy. byx is [0, 1]

7. The two regression lines with variables x and y interest at ( x  y ).

Regression
*206* Solution Manual to Basic Mathematics

8. If the correlation co-efficient rxy is positive then slope of the regression line is positive.
9. The regression co-efficient are independent of change in origin but not scale.
10. In regression analysis, both variables are one is independent and other is dependent.
11. In regression line there is cause and effect relationship.
1
12. If bxy = 6 and byx= , then bxy. byx = 1.5  [0, 1] so it is not possible
4
13. The given lines can be rewrite as
2
First line: x = y + 4 (x on y)
3
Second line: y = 6x + 8 (y on x)
2
 bxy = and byx = 6
3
2
Here bxy = × 6 = 4 > 1 not possible so, again we rewrite the given lines
3
3
First line: y = x – 6 (y on x)
2
1 4
Second line: x = y + (x on y)
6 3
3 1
 byx = and bxy =
2 6
1 3 1
Here, bxy. byx = . =  [0, 1]
6 2 4
1 1
 rxy = bxy. byx = = = 0.5
4 2

14. Since two lines of regressions intersecting at ( x  y ) , we have

5 x – 4 y = 40 ……. (i)

4 x – 3 y = 35 …… . (ii)

elimination x from (i) and (ii)

20 x – 16 y = 160

20 x – 15 y = 175

– + –

– y = – 15  y = 15

Substituting y = 15 in (i)

5 x – 4 × 15 = 40  x = 20

 mean of x - series( x ) = 20
mean of y-series ( y ) = 15

Statistics
 Statistics and Probability* 207*

15. 3x + 5y = 15
3
 y=– x + 15, which is regression line y on x.
5
–3
 byx =
5
16. The regression equation x on y is
3x – y = 20 …… (i)
Here, independent variable (y) = 12, then from (i),
4x – 12 = 20 x=8
 dependent variable (x) = 8
17. In regression analysis, both variable must have same units or different units.
18. Regression lines
2x + 75 = 15 ………. (i)
3x + 5y = 18 ………. (ii)
To find x when y = 3, first we determine the line x on y.
So, given equation can be rewrite so that product of their slopes lie in [0, 1]
2 15
Line (i): y = – x+ [ y on x]
7 7
5
Line (ii) : x = – y+6 [x on y]
3
2 5 10
Obviously, byx × bxy = –  – =  [0, 1]
7 3 23
 Using the line x on y, when y = 3
5
x=– × 3 + 6 = 1.
3

Regression