INTRODUCTION TO BUSINESS STATISTICS

Skewness and Kurtosis

Skewness is a measure of symmetry of data it gives us the pictorial distribution of data. There
are three types of distribution namely
(i) .Symmetrical distribution
(ii) Positively skewed distribution
(iii) Negatively skewed distribution

Coefficient of skewness
skp

Karl Pearson Coefficient of skewness (Skp)

Interpretation
i)
Skp > 0 the data is positively skewed
ii)

Skp < 0 the data is negatively skewed

iii)

Skp = 0 the data is symmetrical

mean mod e
s.deviation

Kurtosis
-

The degree of peakedness of the distribution

It tells how high a mean value is from maximum

To calculate the coefficient of kurtosis we apply the moments

2
Moment coefficient of Kurtosis

m4
m22

In general the rth moment of observation x1,x2,x3........xn about the mean

1 n
r
Mr xi x
r 0,1,2.......n
n i 1
First moment about origin 0

1 NELSON KIPRONO BII Statistics Class notes

is given by

M1

1 n
xi - 0 1

n i 1

1 n
xi x
n i 1

2nd moment about the mean

1 n
2
M 2 xi x variance (x)
n i 1
Example 1
Find the coefficient of Kurtosis from the following data
X
F

0
1

1
2

2
3

3
5

4
6

5
7

6
4

7
3

8
2

9
1

10
0

Solution
m
2 42
m2
Ungrouped data
1 n
4
M 4 xi x
n i 1
M2

1 n
xi x 2

n i 1

Grouped data
1 n
4
M 4 fi xi x
n i 1
M2

1 n
2
fi xi x
n i 1

marks

xi x A

xi x A 2

xi x A 4

fi xi x A

-4.47

19.98

399.23

19.98

399.23

-3.47

12.041

144.98

24.1

289.96

-2.47

6.1009

37.22

18.3

111.66

-1.47

2.1609

4.67

10.80

23.35

15

-0.47

0.2209

0.049

1.33

0.294

24

0.53

2.34

5.47

15.12

32.69

35

1.53

6.10

37.22

24.4

148.88

24

2 NELSON KIPRONO BII Statistics Class notes

fi xi x A

fx

2.53

12.04

144.98

36.12

434.94

21

3.53

19.98

399.20

39.96

798.4

16

4.53

29.92

895.21

29.90

895.21

10

5.53

41.86

1752.33

f 34

fx 152

fx
f

152
4.47
34

1 n
4
fi xi x

n i 1

M4

fi xi x

3134.614

fi xi x

1
3134.614
(3134.614)
n
34
92.19

1
220.03
34
6.47

1 n
2
M 2 fi xi x
n i 1

Interpretation of coefficient of Kurtosis,

m4 92.19

6.47
m22

If the coefficient of kurtosis is such that

2 3

2 3

220.03

i 1

i 1

2 3

then distribution is Normal (Mesokurtic distribution)

then distribution is Leptokurtic (highly peaked)
then distribution is Platykurtic (flatter distribution)
Self- practice questions.

3 NELSON KIPRONO BII Statistics Class notes

Q1. The first four moments of a distribution about

x2

are 1, 2.5, 5.5 and 16. Calculate the

four moments about

and about zero.

Q2. The first four central moments of distribution are 0, 2.5, 0.7 and 18.75. Comment on the
skewness and kurtosis of the distribution
Q3. The daily expenditure in $ of 100 families is given below
Daily
0-20
expenditur
e
Number of 13
families

20-40
f2

40-60
27

60-80
f4

80-100
16

If the mode of the distribution is 44, find the missing frequencies hence calculate Karl
Pearsons coefficient of skewness.
Self-practice questions.
Q1.

In a factory manufacturing pens, machines X, Y and Z manufacture 30, 30 and 40 per

cent of the total production of pens, respectively. Of their output 4, 5 and 10 percent
of the pens are defective. If one pen is selected at random, and it is found to be
defective, find from which machine the pen comes.
(Z=0.6639)

Q2.

Suppose 80 per cent of the material received from a vendor is of exceptional quality,
while only 50 per cent of the material received from vendor B is of exceptional
quality. However, the manufacturing capacity of vendor A is limited and for this
reason only 40 per cent of the material purchased comes from vendor A. The other 60
per cent comes from vendor B. An incoming shipment of material is inspected and it
is inspected and it is found to be of exceptional quality. What is the probability that it
came from vendor A.

REGRESSION ANALYSIS
Definition: regression is the measure of average relationship between two moving variables
in terms of origin unit of data.
Use of regression/ importance
It provides estimate of value of dependent variable from value of independent variable.
used to obtain a measure of error involved in using the regression line for estimation
with help of regression analysis we can obtain a measure of degree of association of
correlation that exist between two variables
Regression equation
4 NELSON KIPRONO BII Statistics Class notes

For two variables x and y we can have the regression equation

i) Regression of y on x

y f (x)

x is independent variable and y is dependent variable

x f ( y)

ii) Regression of x on y
y is independent variable and x is dependent variable
Note: in regression analysis the independent variable is also known as regressor or predictor
or explanatory and dependent variable is called regressed or explained variable.
Method of least Square
y a bx
a, b are constants
to solve the equation we use the following relationship
y Na bx
xy ax bx 2

these are NORMAL equations

x on y
x Na by
xy ay by 2

Example 1
Determine regression equation of y on x from the following data
X
Y

1
2

2
5

3 4
3 8

5
7

Solution
X
1

Y
2

X2
1

Y2
4

XY
2

25

10

16

64

32

25

49

35

X 15

Y 25

X 2 55

Y 2 151

XY 88

(i)
Regression of X on Y
x a by

x Na by
xy ay by 2

5 NELSON KIPRONO BII Statistics Class notes

15 5a 25b
88 25a 151b
75 25a 125b
88 25a 151b
13 26b
b 0.5

15 5a 12.5
2 .5 5 a
a 0.5

y a bx

y Na bx
xy ax bx 2
25 5a 15b
88 15a 55b
75 15a 45b
88 15a 55b
13 10b
b 1.3
a 1.1

Example 2
(ii) Regression of Y on X

X 1.1 1.3 y

6 NELSON KIPRONO BII Statistics Class notes

A company is introducing a job evaluation scheme in which all jobs are graded by points for
skill, responsibility and so on. Monthly pay scales (in thousand Kenya shillings) are then
drawn up according to the number of points allocated to the number of points allocated. To
date the company has applied this scheme to 9 jobs:
Job
A
B
C
D
E
F
G
H
I
Points
5
25
7
19
10
12
15
28
16
Pay
3.0
5.0
3.25
6.5
5.5
5.6
6.0
7.2
6.1
(a) Fit the least squares regression line for linking pay scales to points
(b) Estimate the monthly pay for a job graded by 20 points

Solution.
Grade
point
5
25
7
19
10
12
15
28
16
137

d x x 15

dx

-10
10
-8
4
-5
-3
0
13
1
2

100
100
64
16
25
9
0
169
1
484

Pay scale,
y

dy y 5

dy

3
5
3.25
6.5
5.5
5.6
6.0
7.2
6.1
48.15

-2.0
0
-1.75
1.5
0.5
0.6
1
2.2
1.1
3.15

4
0
3.06
2.25
0.25
0.36
1
4.84
1.21
16.97

(a)

x 137 / 9 15.22

y 48.15 / 9 5.35
b yx

n d x d y d x d y
n d x d x
2

9(65.40) 2(3.15)
9( 484) ( 2) 2

=0.133

Substituting in

y y b yx ( x x)

We have y-5.35=0.133(x-15.22)
=3.326+0.133x
(b) For job grade point 20, the estimated average pay scale is given by
3.326+0.133(20)
=5.989 x 1000
=Kshs 5989

dxd y
20
0
14
6
-2.5
-1.8
0
28.6
1.1
65.40

Self practice question.

The following table shows the score made by sales men on intelligence test and weekly sale
Sales man
Test score
sales

1
40
2.5

2
70
6.0

3
50
4.0

4
60
5.0

5
80
4.0

6
50
2.5

7
90
5.5

8
40
3.0

9
60
4.5

10
60
3.0

Calculate the regression line of sales on test scores and estimate the probable weekly sales
volume if a sales man makes a score of 100.
Activity: Refer to reading texts and read on correlation analysis
then answer the following questions;
Q1. Define the term correlation
Q2. Explain the different types of correlation.
Q3. Explain areas of application for correlation analysis in
business
Q4. Using the self practice question above, find the correlation
between test score and sales, hence comment on your results.
Q5. The following data give the ages and blood pressure of 10 women
Age
Blood
pressure

56
147

42
12
5

36
118

47
128

(a)
Find the correlation
(Answer=0.892)

49
145

42
140

coefficient

60
15
5

72
160

between

age

63
149

55
15
0

and

blood

pressure

(b) Determine the least squares regression equation of blood pressure on age
(Answer, Y=83.758 +
1.11x)
(c) Estimate the blood pressure of a woman whose age is 45 years (Answer=134)