ABM 401 Lesson 11

Lesson 11
Measures of Skewness

INTRODUCTION

Skewness is used in the opposite sense of symmetry. It measures the degree of

distortion from symmetry. In a frequency distribution, skewness is the lack of
symmetry and indicates lopsidedness of the curve. If i n a distribution, the largest
category of items does not occur at the centre of the distribution, but drifts to the
left or to the right, then it is called a skewed distribution. When a distribution is
symmetrical, the values of the Mean, Mode and Median coincide. Skewness has
the effect of pulling the median and mean away from the mode, sometimes to the
right and some-times to the left.

According to Simpson and Kafke, “Skewness or asymmetry is the attribute of a

frequency distribution that extends further on one side of the class with the highest
frequency that on the other”. Garrett defined it as “a distribution is said to be
skewed when the mean and the median fall at different points in the distribution,
and the centre of gravity is shifted to one side or the other – to left or right”.
Similarly, Croxton and Cowden viewed that “When a series is not symmetrical, it is
said to be asymmetrical or skewed”.

Some important points regarding distribution of items in a series and skewness are:
In a symmetrical distribution, frequencies on the either side of Mode are
equal, and the values of Mean, Median and Mode are alike or equal.
In a symmetrical distribution, First and Third Quartile are equi -distant. It
means (Q3-M) = (M-Q1).
In a symmetrical distribution, Sum of positive deviations is equal to Sum of
negative deviations.
In a positively skewed distribution the value of mean is greater than the mode.
The value of median would be less than the mean but greater than the mode.
It means Mean> Median> Mode.

146
ABM 401 Lesson 11

In a negatively skewed distribution mode is maximum and the value of

arithmetic mean is the least. The value of median is less than the mode but
greater than the mean. It means Mode> Median> Mean.
It should be noted that in moderately symmetrical distributions, the interval
between the mean and median is approximately one third of the interval
between the mean and the mode.
Tests of Skewness:

Several tests may be applied to know the presence of skewness in a given series.
However a distribution is said to be skewed if:
The values of mean, median and mode are not same.
The lower quartile and the upper quartile are not equi-distant from the median.
The algebraic sum of deviations of items from the mean is not equal to zero.
Different tests may sometimes give different results according to different angles of
study.

Measures of Skewness:

Measures of skewness tell us the direction and extent of asymmetry in a series,

and permit us to compare two or more series. These measures can be either
absolute or relative.

(i) Absolute Measures of Skewness: Skewness can be measured in absolute

terms by taking the difference between mean and mode. Symbolically,
Absolute Skewness = Arithmetic Mean - Mode
If the value of mean is greater than mode, skewness will be positive.
Conversely, if the value of mode is greater than mean, skewness will be
negative.
(i) Relative Measures of Skewness: In relative measure of skewness, the
difference between mean and mode is placed in relation to standard deviation
of the distribution. There are four important measures of relative skewness,
namely, Karl Pearson's coefficient of Skewness; Bowley's coefficient of

147
ABM 401 Lesson 11

Skewness; The Kelly's coefficient of Skewness, and Measure of skewness

based on moments. In this lesson we will discuss only first three measures.

Karl Pearson's coefficient of Skewness (popularly known as Pearsonian coefficient

of skewness), is based on the difference between mean and mode. According to
his formula of relative measure, the difference between mean and mode is divided
by standard deviation. The coefficient of Skewness calculated by Karl Pearson’s
formula usually lies between –1 to +1. When a distribution is symmetrical the
values of mean, median and mode coincide and therefore, the coefficient of
skewness will be zero. But, when distribution is positively skewed, the coefficient of
skewness shall have plus sign. Similarly, when it is negatively skewed, the
coefficient of skewness shall have minus sign.

FORMULAE FOR CALCULATING SKEWNESS AND

COEFFICIENT OF SKEWNESS
Formula
Measure
Skewness Coefficient of Skewness
Karl Pearson’s
Measure
(First Measure of
Skewness) When Z is ill defined.

Bowley’s Measure
(Second Measure Or Or
of Skewness)

Kelly’s Measure Or Or

Bowley's Coefficient of Skewness is based on quartiles. It indi cates that in a

symmetrical distribution first and third quartiles are equi-distant from the median,
so, coefficient of skewness will be zero. In positively skewed distribution, it is
positive and in case of negatively skewed distribution it is negative. Kelly's

148
ABM 401 Lesson 11

Coefficient of Skewness measures skewness on the basis of 10 th and the 90 th

percentiles (or the first and ninth deciles).

Example 1: From the observations 36, 32, 41, 38, 33, 37, 30, 35, 39, 35 calculate
Karl Pearson’s Coefficient of Skewness.

Solution:
X 36 32 41 40 38 33 37 30 35 39 35
d 0 -4 5 4 2 -3 1 -6 -1 3 -1
0 16 25 16 4 9 1 36 1 9 1

Mode (Z) by Observation 35.

Example 2: From the distribution given below, find out: (i) Karl Pearson’s Coeff. of
Skewness; (ii) Quartile Coeff. of Skewness.
Ht. in CM. 75 76 77 78 79 80 81 82 83
No. of Student 6 8 13 18 20 16 10 7 2
Solution:

75 6 -4 -24 96 6
76 8 -3 -24 72 14
77 13 -2 -26 52 27
78 18 -1 -18 18 45
79 20 0 0 0 65
80 16 1 16 16 81
81 10 2 20 40 91
82 7 3 21 63 98
83 2 4 8 32 100

149
ABM 401 Lesson 11

(i) Karl Pearson’s Coefficient of Skewness:

(ii) Quartiles Coefficient of Skewness:

Example 3: Find the coefficient of skewness of the groups given below and point
out which distribution is more skewed.

Marks Group A Group B

55-58 12 20
58-61 17 22
61-64 23 25
64-67 18 13
67-70 11 7

150
ABM 401 Lesson 11

Solution:
GROUP A
MV (X)
55-58 56.5 12 -2 -24 48 12
58-61 59.5 17 -1 -17 17 29
61-64 62.5 23 0 0 0 52
64-67 65.5 18 1 18 18 70
67-70 68.5 11 2 22 44 81
N=81

GROUP B
MV (X)
55-58 56.5 20 -2 -40 80 20
58-61 59.5 22 -1 -22 22 42
61-64 62.5 25 0 0 0 67
64-67 65.5 13 1 13 13 80
67-70 68.5 7 2 14 28 87
N = 87

151
ABM 401 Lesson 11

Example 4: If the Median and Mode of any 10 item values are 25.5 and 28
respectively and the sum of squared deviations from mean is 360, find out Karl
Pearson’s Coefficient of Skewness.
Solution: Given- N=10, Median=25.5, Mode=28, and =360

Mode = 3 Median - 2 Mean

28 = (3x25.5) – (2xMean)

28 - 76.5 = 2 Mean

So, Mean = 24.25

SUMMARY

Skewness is the lack of symmetry and indicates lopsidedness of the curve. If in a

distribution, the largest category of items does not occur at the centre of the
distribution, but drifts to the left or to the right, then it is called a skewed
distribution. In a skewed distribution: values of mean, median and mode are not
same; the lower quartile and the upper quartile are not equi-distant from the
median; and the algebraic sum of deviations of items from the mean is not equal to
zero. There are two measures of skewness, viz. absolute or relative. These
measures can be calculated by methods given by Karl Pearson, Bowley, and Kelly.

152
ABM 401 Lesson 11

REVIEW QUESTIONS
1. What do you understand by positive and negative skewness? Explain the
objectives and various methods of measuring it.
2. Distinguish between dispersion and skewness. What are the tests of
skewness?
3. Define Skewness? When is a curve said to be skew? How is the skewness of
the curve measured?
4. Calculate Quartile Coefficient of skewness of the following observations:
15, 21, 18, 17, 18, 21, 10, 12, 16, 14, 11, 22, 9, 16, 18. Ans. = -0.33
5. From the data given below find out as to which group is more skewed? Use
mean and median.
Group A 100 118 122 109 105 107 121 113 105 100
Group B 18 26 21 25 23 22
15 16 17 17
Ans: = 0.78, 0.40
6. Compute quartile measure of skewness of the following data:

Size 1 2 3 4 5 6 7 8 9 10
Frequency 2 9 11 14 20 24 20
5 16
2
Ans. = -0.33
7. Calculate Karl Pearson’s Coefficient of Skewness from the following data:

Marks 15 18 20 24 27 29 30
Frequency 11 15 22 17 8 10 7
Ans. = 0.458
8. Calculate the Karl Person’s Coefficient of Skewness from the following table:
Marks 70-80 80-90 90-100 100-110 110-120 120-130 130-140 140-150
Freq. 6 9 17 21 25 23 10 8
Ans. = - 0.27

SUGGESTED READINGS
Elhance DN: Fundamentals of Statistics
Gupta SP: Statistical Methods
Gupta BN: Statistics
Nagar KN: Fundamentals of Statistics
Varshney RD: Fundamentals of Statistics
Nagar AL: Fundamentals of Statistics

153